U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
|Publication Number: FHWA-HRT-11-064 Date: November 2011|
Publication Number: FHWA-HRT-11-064
Date: November 2011
This chapter provides guidance on the selection of traffic analysis tools, parameters, and assumptions to support consistent analysis throughout the project development process. This chapter builds on information about the strengths and weaknesses of each tool category that can be found in volume II of FHWA's Traffic Analysis Tools Program.(1) The general traffic analysis tool types and their salient strengths and weaknesses were covered in chapter 2. The MOEs output by each tool type were covered in chapter 4.
Traffic analysis tools have been developed and designed to address each stage of the project development life cycle. Some tools are useful for several stages while others are best limited to a single stage of the process (see figure 1).
As shown in figure 1, each tool category has both its feasible range of application and its optimal range of application within the project development stage. In many cases, multiple tool categories can be applied at the same project development stage. However, as discussed in chapter 4, it is best to use the tool best designed for each stage of the process at the appropriate stage.
Note that while the ranges of application overlap significantly, no single tool will carry the analyst through the entire project development process. Similarly, the analyst often has more than one tool type to choose from and may employ more than one tool type at each stage of the project development process.
Details of the tool categories and their uses are as follows:
At the project need and initiation stages, one is concerned with identifying system problems and possible solutions. A wide geographic coverage and the ability to easily test several widely varying alternative solutions are desirable.
Little is known about the project at this time, so it is not possible to model the facility operations in detail. Thus, the efforts necessary to conduct operations analyses using HCM analysis tools, optimization tools, and simulation models will not produce the desired precision of the results. The approximate representation of facility operations in the demand model is sufficient at this stage of the analysis to identify future problem spots and possible alternative solutions.
Study limits, forecast years, parameters, and assumptions should be kept conservative to reflect uncertainty at these early stages of project development.
The travel demand model is the best tool type to apply at these early stages of analysis. Sketch planning models may be used to supplement the demand model for conditions where the demand models lack the appropriate sensitivity (e.g., testing congestion pricing in a region where there are no existing toll facilities, and consequently, the demand model is not sensitive to tolls).
If the demand model already incorporates all of the desired sensitivities for evaluating project alternatives, then there is usually little or no need to employ sketch planning. Sketch planning should be used to introduce variable sensitivities not otherwise already incorporated in the demand model.
Both sketch planning and travel demand models will generate systemwide MOEs such as VMT and VHT. However, the greater network detail of the typical demand model suggests that the travel demand model will produce more reliable estimates of systemwide performance. The sketch planning model results should be used to modify the demand model systemwide results, not replace them. For example, if the sketch planning model predicts that congestion pricing will change systemwide VMT by 5 percent, the demand model VMT should be changed by 5 percent to reflect the effects of congestion pricing.
The forecast years should be conservatively far into the future to allow for slippages in the project delivery schedule. Thus, if the requirement during the project clearance stage is to evaluate project impacts for 20 years after project opening, the manager may elect to develop a 25- or 30-year forecast for the early stages of the project development process.
The temporal (time of day) and geographic limits of the study area should be oversized during the early stages of the project development process to ensure that the limits of the actual project impacts can be identified early in the process and result in significant savings in analysis effort during the later stages. A useful analysis at the project initiation stage is to identify the likely geographic and temporal bounds of the project impacts on traffic operations of nearby facilities. This will help scope the analysis during the project clearance stage.
Cumulative Improvement Assumptions
Generally, one should use the highway and transit network improvement assumptions incorporated in the agency's currently adopted long-range transportation plan. The manager will have to select between the agency's target or desired improvement program and its financially constrained improvement program. The selection will be based on the manager's experience as to decisionmakers' (those that will be involved in later stages of the process) perceptions of the relative acceptability of the different improvement programs for use in the environmental analyses.
Capacities should be conservatively low during the early stages of project development. For example, level of service C capacities (approximately 80 percent of true capacity) may be used instead of HCM capacities in the planning analyses. This builds in a cushion for uncertainties in assumptions, project design, and operation at the early stages of project development.
The project clearance stage is the most challenging stage for the manager to select and coordinate traffic analysis tools, MOEs, and assumptions. This is because this stage requires the most analysis and because there is a wide range of tools to accomplish the analysis.
The manager cannot avoid applying multiple tools at this stage. Project clearance requires detailed analyses of the proposed project demands and operations. Project clearance also requires a more far-ranging environmental analysis of the project and feasible alternatives. The environmental analysis requires a systemwide evaluation of the project effects. Thus, multiple tools are required, and the manager and analyst must take steps to ensure that the tools work together consistently to produce valid results.
The two keys to managing consistency in the project clearance stage are as follows:
Using each tool only for the analyses it is best at means, for example, not using a travel demand model to report specific facility operating results when a similar analysis is available from a superior operations analysis tool such as the HCM, optimization, or simulation models.
Scoping broadly and conservatively means that, early in the project clearance stage, the manager and analyst should adopt relatively large geographic and temporal boundaries for the analysis until the alternatives and the project impacts are better defined and the scope of analysis can be narrowed.
The analysis requirements of the project clearance stage include demand forecasting and performance analysis. Both of these analyses need to be done at the systemwide and facility-specific levels.
The long-term systemwide and facility-specific forecasting of demands are best done using a travel demand forecasting model. This is the only available tool type for these analyses. Sketch planning tools generally do not provide sufficient details for the project clearance stage. If sketch planning models are used, it should be to fill one or more gaps in the sensitivities of the travel demand forecasting model to features of the project or its alternatives.
Caution should be used when using demand models to forecast specific facility conditions in the short term (under 5 years). Residual calibration error for the facility may be greater than the forecasted growth for the facility. In such cases, the raw model forecasts should be adjusted for residual validation year error for the subject facility. This will ensure that the short-term model forecasts for the facility are consistent with the existing year counts.
In all cases, when using a demand model to forecast facility-specific demands, the manager and analyst should employ the following steps to ensure consistency:
Adjust demand model forecasts for differences between facility counts and count year model estimates. These become the facility segment-specific volume corrections that should be added to the model forecasts for the facility. Alternatively, the analyst may use the model to develop growth factors that are applied to existing counts. Either way, the analyst should use and report the adjusted forecast results rather than the raw model forecasts for the facility.
Validate that the forecasted demand model operations for the facility are similar to those predicted by an HCM-type tool (or optimization tool or simulation model) for the facility. If the overall average speed for the facility is within tolerances for both the demand model and the HCM-type (or other operations) tool, then the analyst can be confident that the demand model is predicting approximately the right demands for the facility. Greater differences should be corrected by examining and adjusting the demand model capacities, free-flow speeds, and volume-delay functions for the facility. See Travel Model Validation and Reasonableness Checking.(6)
When using a demand model to determine the geographic scope of project demand effects, be sure to use a sufficient number of equilibrium iterations and a tight enough equilibrium closure criterion to ensure that forecasted project/no-project differences in link volumes are truly due to project differences and not to differences in how close the model got to equilibrium each time. Volume effects that are not contiguous (a minor volume change in a subarea with no other volume changes) should be considered equilibrium closure inaccuracies rather than project effects.
DTA is the loading of traffic onto a highway network in a sequential series of time slices within the overall analysis period (usually the peak period) using a procedure called dynamic user equilibrium (DUE). DUE is the equilibration of traffic based on experienced travel times.(7) As agencies and their personnel build up experience with DTA, the manager may consider employing this tool to improve the accuracy of the facility performance predicted by the demand model. If the agency has little experience applying DTA, the manager should be cautious about ensuring sufficient contingencies are built into the budget and schedule to overcome unexpected difficulties that can arise when using a new tool for project delivery purposes.
If DTA has not been previously employed with the demand model, the manager should realize that a complete recalibration and revalidation of the demand model will be necessary. DTA will throw off the previous calibration and validation of the demand model.
System performance analysis (e.g., VHT, VHD, etc.) can be done using either travel demand models or simulation models. HCM-type tools are not currently capable of system performance analysis.
Using an existing travel demand model will be much less resource-intensive than developing a simulation model from scratch. However, the demand model will produce less precise performance results. Overall system performance may be about right in the demand model, but the specific links and facilities that the demand model predicts are congested may not be right.
Exceptionally large study areas are most cost-effectively modeled in travel demand models. Practical resource considerations limit the attractiveness of using simulation models for study areas over 20 mi in length. Simulation can be and has been done for larger areas, but the resource requirements (staff, data, and time) are high.
Improving the System Performance Accuracy of Demand Models
Concerns about the accuracy of the systemwide performance predictions from demand models can be controlled and managed as follows:
Emerging Option—Adding DTA
DTA shows promise for improving the system performance predictions of travel demand models without incurring the great expense of creating a simulation model. DTA may also facilitate multiresolution modeling by achieving better consistency between travel demand models and HCM/simulation models. However, if an agency has little or no experience with DTA, it should be treated as a research effort (with the inherent schedule and budget risks associated with research) until the agency acquires more experience with it.
Use of Simulation Models for Systemwide Performance Analysis
For moderate- to small-sized study areas, study resources may permit the development and use of a simulation model to predict system performance in lieu of the travel demand model. A simulation model will provide much superior system performance results for existing and near-term conditions as long as congestion can be kept within the geographic and temporal bounds of the simulation model. Once these bounds are exceeded, the simulation model may provide system performance results that are inferior to that of the demand model. The simulation model may still more accurately predict the performance of specific segments within the facility, but its estimates of systemwide delay will be thrown off by congestion spilling over the borders of the model.
If a simulation model is to be used to produce system performance results, a demand model will still be needed to produce the demand forecasts. The manager and analyst must ensure that there is approximate consistency between the system performance information used by the demand model to predict demands and the system performance predicted by the simulation model for those predicted demands.
It is suggested that the predicted systemwide average speeds of the two model types (demand and simulation) be compared to ensure that they are approximately consistent.(6) This check should be made for both existing and future conditions. The analyst should not force the demand model to predict the same locations and amounts of congestion as the simulation model on a link-by-link basis (although this kind of examination may be useful in understanding sources of discrepancies and fixing them). The goal is to ensure the demands produced by the demand model are roughly in agreement with the performance predicted by the simulation model.
The analysis of facility-specific performance for future conditions is best done using HCM-type tools, control optimization tools, and simulation models. Control optimization tools have built-in simulators that are excellent for forecasting facility performance under prevailing control conditions. HCM-type tools are excellent for sizing a facility to avoid bottlenecks. Simulation models are best for situations where a simpler HCM-type or optimization tool cannot produce the needed performance accuracy. Simulation models are also better than HCM-type tools for evaluating oversaturated conditions, in which demand is greater than capacity.
Demand models are generally not satisfactory for evaluating facility-specific performance. The peak spreading effects of upstream bottlenecks on downstream volumes are not treated in conventional travel demand models. Congested speeds may be shown for the wrong segments of the facility. Demand models typically show congestion occurring on the bottleneck segment, rather than upstream of the bottleneck. DTA can improve this capability, but the manager should ensure that his or her agency has or can obtain the necessary expertise before employing it in project development work.
The manager's first choice for analysis tools at the individual facility level should be the optimization tools. If an optimization tool is available for the selected facility type, then the optimization tool combines the relative data efficiency of HCM-type analyses with the ability to optimize control settings during the operation of the facility.
Should a suitable optimization tool not be available, the next best choice is an HCM-type tool. This type of tool has a relatively high data efficiency (least data required to produce suitably reliable performance results).
Should the project being evaluated involve sophisticated operations management features not incorporated in the HCM methods or should it involve dealing with the management of congestion spillovers within the facility, then the manager should use a simulation model. Macroscopic or mesoscopic simulation models are more data-efficient then microscopic simulation models. However, if the HCM-type tools are unable to deal accurately with the situation, it will probably be a challenge for the macroscopic models and may be a challenge for the mesoscopic simulation models.
At the current state of the practice, the manager and analyst are usually confronted with developing a microsimulation model from scratch. This is often as resource-intensive as creating a demand model from scratch. Should a previously coded and calibrated DTA model be available for the project site, then some data-collection effort for the microsimulation model may be avoided. This has been the basis of the discussion regarding the selection of analysis models; however, should one be fortunate enough to already have a calibrated simulation model available for the facility, the data requirements implicit in the previous discussions are avoided and the use of a previously calibrated microsimulation model can be more cost-effective than even HCM analyses.
Managing Consistency at the Facility-Specific Level
Whichever tool is selected to perform the facility-specific performance analysis, that tool should be used to report all facility performance results. For example, if CORSIM was used to generate the facility travel times, it should also be used to identify bottlenecks, queues, flows, average speeds, and delays for the facility. Unless the manager has some doubts about the tool, there is little reason to use multiple tools to do the same thing. The best way to avoid consistency problems is to avoid unnecessarily introducing inconsistency.
The facility-specific results from travel demand models should be checked for general consistency with the selected traffic operations analysis tool (HCM, optimization, simulation), as previously explained. However, the facility-specific results from demand models should generally not be reported. DTA may eventually overcome this weakness of travel demand models for predicting facility segment-specific performance.
The forecast years need not be as far into the future as they were in the earlier project delivery stages; however, significant delivery schedule slippages can still occur even within the project clearance stage. Consequently, it pays to be a bit conservative to allow for possible slippages in the project delivery schedule. Thus, if the requirement during the project clearance stage is to evaluate project impacts for 20 years after project opening, the manager may elect to develop both a 20-year forecast and a 25-year forecast just in case there are unexpected delays and slippages during the environmental review process.
The temporal (time of day) and geographic limits of the study area should be better bounded at this stage to avoid over-expenditure of resources during the project clearance process. The manager and analyst may elect to slightly oversize the limits so decisionmakers can see that all relevant impacts have been covered. Alternatively, the manager may point to earlier analyses that bounded the impacts of the project during the project initiation stage.
Cumulative Improvement Assumptions
The manager should recognize that, at the project clearance stage, decisionmakers and stakeholders will have a significant influence on the cumulative improvement assumptions to be made for the project analysis. An early meeting with these participants to pin down and document assumptions may be very effective. If such a meeting cannot be practically scheduled before the analysis must be performed, the manager may attempt to bracket the potential input from decisionmakers by selecting a few sets of assumptions and splitting them into high-, medium-, and low-level scenarios. The analysis is then repeated for each scenario of assumptions.
Capacities used during the early and later steps of the project clearance stage should generally be as consistent as possible, recognizing that they will invariably change as the project design and its alternatives are refined. If the analyst is confronted with a great deal of uncertainty regarding the project design at the start of the project clearance stage, he or she might resort to high-, medium-, and low-level scenarios for the project design and then discard the irrelevant scenarios as the project is better defined.
The objectives during PS&E, construction, and operation are the design, construction, and operation of the facility. During these stages, the project specifics are defined, built, and operated. The focus is on specific bottlenecks in the facility and the overall facility operation. At these stages of project development, HCM-type tools, control optimization tools, and simulation models are the best available tools for assessing facility and bottleneck performance.
The HCM-type tools are best for sizing and designing facilities to avoid congestion. They are quick to apply and require comparatively little data. Calibration data is generally not required. The rapid-response capabilities of HCM-type tools make them useful for design and construction. However, their lack of optimization capabilities makes HCM-type tools less useful for operating a facility.
Optimization tools generally automate the evaluation of multiple bottlenecks within a facility better than the HCM. However, most optimization tools, being based on HCM methods, suffer the same weaknesses as the HCM when dealing with the spillover effects of congestion on facility operation. Optimization tools are designed for operating the facility.
Simulation models should be used where constraints do not permit sizing the facility to avoid congestion and it is necessary to manage the spillover effects of congestion within the facility. Simulation models are the most resource-intensive of the three tool types discussed. They must always be calibrated. The resource and time requirements generally make the creation of new simulation models from scratch of less use during construction of the facility. Their lack of optimization capabilities generally reduces the value of simulation models for facility operation.
During the later stages of project development, the greatest concern is delivering a project that will be successful in its real-world setting. Consistency in these later stages is a good first-order quality control check on the analysis. However, being right is more important than consistency in these later stages of project development.
This case study is a continuation of HOV/HOT lanes project described in previous chapters. The manager is in the process of developing the overall PDAP and has identified the desired MOEs for the various stages of the project. Now, the manager must identify the proposed analysis tools for each stage and the process that will be used to ensure consistency among the various tools. Table 6 provides an overview of the selected tools by stage of project development.
Project Development Stage
System and facility MOEs
Travel demand model supplemented with sketch planning model
Sketch planning model used to help demand model estimate demand effects of tolls for HOT lane alternative. Facility results used to help scope project, its alternatives, and likely scope of project effects on facility demands. Numerical performance results not considered definitive for a specific facility.
Travel demand model supplemented with sketch planning model
Models used to estimate system and facility demands. Models used to estimate performance only for system MOEs.
HCM (or optimization model) supplemented with microsimulation
The selected HCM (or optimization) tool will predict mixed-flow lane operations but is unable to deal with proposed HOT lane access options. Microsimulation will be used to evaluate a few prototypical access options, and the results will supplement the HCM results.
PS&E, construction, and operation
System effects were dealt with in previous stage.
HCM or optimization tools
Simulation models would be used only if available from earlier stage.
Notes: The selected system MOEs are VHD, average speed, mode split, and VMT by speed bin. The selected facility MOEs are delay/vehicle, speed, level of service, HOV volumes, transit patronage, collision rate during construction, and collision rate after open.
During the project needs and initiation stages of project development, the focus is necessarily on a large possible project impact area and a wide-ranging set of alternatives. In these early stages of the HOV/HOT lanes project development, a previously calibrated and validated regional or subregional travel demand is the best tool for analyses.
Selection of Primary Analysis Tool
In this case study, two previously calibrated demand models were considered by the manager: the regional travel demand model and a subregional, countywide model. Both were conventional four-step models with trip generation, distribution, mode choice, and traffic/transit assignment capabilities. The county model was designed to be consistent with the regional model with additional network and land use detail within the county.
Because of the superior network and land use detail in the vicinity of the proposed HOV/HOT lane project and the superior external trip forecasting capability, the county model should be selected as the demand model to be used throughout the project development process.
Selection of Supporting Tool
The demand model should be satisfactory for predicting the requisite system MOEs for the early stages of project development: VHD, average speed, and mode split. However, it is not be sufficient to predict the effects of the project alternative of installing an HOT lane instead of an HOV lane.
In this case study, the county model, like the regional model, has the capability to test the effects of tolls on mode choice but lacks the ability to test the effects of tolls on route choice (facility choice). Consequently, a sketch planning model is needed to identify how many of the county model-predicted single-occupancy vehicles (SOVs) on the freeway will be willing to pay a toll to use the HOT lanes. The demand model is used to predict how many SOVs and HOVs will use the total freeway facility. The sketch planning model will be used to split the SOVs using the freeway into toll-paying and non-toll-paying SOVs.
The manager would investigate the available toll analysis sketch planning tools used in previous HOT lane studies. In this case study, the toll elasticities in the regional model would be considered sufficient to construct a simple spreadsheet post-processor. For each cell in the freeway facility origin-destination table, the spreadsheet reads the relative travel times in the mixed-flow lanes and the HOT lanes, applies the toll elasticities, and predicts the percent of SOVs that will pay the toll for different toll schedules.
Selection of Assumptions
Toll schedule: At this early stage in project development, the toll schedule would not be known, so the manager and analyst assume that the toll schedule would be set to keep the total volume in the HOT lane at or near a target of 1,650 vehicles per hour (a target set separately by the agency operating the HOT lanes). The sketch planning tool, the SOV toll processor, would then be used during the project clearance stage to compute the likely toll schedule required to approximately achieve that target.
HOV operating hours and occupancy: Ordinarily, the manager would be able to assume the standard agency policy for HOV operating hours and occupancy requirements. However, the HOT lane alternative introduces more complexity. Thus, the manager should be prepared to scope the study to address multiple HOV occupancy requirements and the possibility of extended or all-day operation.
Forecast years: The manager would look at the available forecast year data sets in the selected travel demand model and pick the forecast year that falls closest to one of the available model data sets. If necessary, the manager may scope for linear extrapolation of the model forecast data set to a later year to provide sufficient cushion for slippages in the project delivery schedule. In this case, a 2030 forecast year was available, but the manager conservatively plans to extend this forecast model dataset to 2035 to allow for delays during the project delivery process.
Assumed cumulative projects: The manager would examine the regional transportation plan and, thinking ahead, select the set of projects that would best meet the needs of the project clearance stage. Since the environmental analysis that is performed during project clearance generally conservatively assumes a financially constrained set of cumulative improvement projects, that is the set the manager would select from the regional transportation plan for this analysis. Ideally, this decision would be made with input from the decisionmakers and stakeholders; however, their input will not be secured until the project clearance stage. So, the manager must make an educated guess based on environmental analyses previously conducted by the agency.
Study limits: The initial study limits are set by the selected travel demand model, the region. The study limits can be refined for later stages of analysis based on the results of these initial studies. The large study area is guaranteed to trap all significant project impacts; however, the manager and analyst should be prepared to deal with a dilution of project effects when using a large study area. If this is an issue, the manager may elect to identify a secondary, focused study area for accumulating system MOEs. This will avoid diluting the project effects. The regional study area is retained to ensure that all significant project effects are trapped in the model, while the secondary, focused subarea is used to accumulate system MOEs that better convey the effects of the project on facilities in the area.
Use of Facility Performance Results
The results of the earlier stages of project analysis will be used to help scope the project, its alternatives, and likely the project effects on facility demands. The numerical performance results for the project facility, which are produced in these early analysis stages, will be replaced later by more definitive results.
During the project clearance stage of project development, the focus is on both project detail and on a wider project impact area. At this stage of the HOV/HOT lanes project development, a combination of system- and facility-specific tools are required to support the analysis.
Selection of Primary and Supporting Analysis Tools—System MOEs
To forecast systemwide and facility-specific demands, the same combination of travel demand and sketch planning models are recommended as identified for the project initiation stage. This combination of demand and sketch planning models should also be used to generate system MOEs.
Selection of Primary and Supporting Analysis Tools—Facility MOEs
To evaluate facility-specific operations, the agency had an in-house tool for evaluating freeway operations. In this instance, the tool was the University of California's FREQ model. The HCM 2010 freeway analysis tool, FREEVAL, could also have been selected; however, with FREEVAL, additional simulation modeling would have been needed to address ramp metering and HOV lane options as well as HOT lane access options. The HCM-based tool used is able to model ramp metering and HOV lanes but was not developed to evaluate the controlled access features of HOT lanes. The manager therefore elected to use a simulation model to test various designs for providing access to the HOT lanes. The simulation targeted answers to HOT lane operations questions that could not be answered with the HCM-based tool.
Since the project will increase the throughput capacity of the freeway facility, it is likely to offload the adjacent surface street system. The demand model may be sufficient to identify the general direction and magnitude of the demand and performance changes for the surface streets.
If the project and one or more of its alternatives were to include a ramp metering component (or other features) that might reduce the throughput capacity of the freeway, then there may be concerns about the impacts on surface streets in the area. Under such a condition, a separate HCM-type tool would be selected to generate performance measures for the adjacent surface street system likely to be impacted by the project. The HCM-produced performance measures for these streets would supersede the performance outputs coming from the demand model. If significant and complex diversion effects are expected, the manager should select a tool that can integrate arterial and freeway performance with a diversion (route choice) prediction capability, such as is provided by many microscopic and mesoscopic simulation models. The ability of the simulation model to accurately represent intersection signal controls and the route choice algorithm employed in the model should be investigated by the manager to ensure that the model will perform as expected.
The project manager would develop an analysis protocol that addresses how freeway demands are handed off to the surface street HCM tool and how surface street demands are handed off to the freeway HCM tool.
The manager should provide for one or more checks to ensure that the freeway and surface street HCM models are working with consistent volumes at their interface in the future forecast years. For example, if the demand model says the demand is X for the off-ramp but the HCM tool says only Y can actually arrive there, the analyst will need to determine if the difference adversely affects the conclusions of the overall analysis. If so, then a simulation model integrating the freeway and surface street system would be required in lieu of the HCM tools.
A simulation model of the corridor would solve the freeway and surface street interface problem by integrating both facility types into a single model. However, creating a simulation model from scratch puts a significant strain on project analysis resources.
Scoping for Facility Closures and Construction Detours
If facility closures are contemplated for the construction phase of the project, it may be highly desirable for the manager to incorporate an HCM-type tool (or optimization or simulation) in the project clearance scope for the nearby off-facility street system to facilitate construction detour planning. Construction detour planning is generally beyond the capabilities of conventional travel demand models.
Selection of Assumptions
Toll schedule, HOV occupancy, hours of operation: At the project clearance stage, the toll schedule would still not be known. However, the environmental analysis will need to address the potential range of tolls that might be used. The manager and analyst would address this issue by positing several toll schedules, HOV occupancy requirements, and hours of operations. Schedules might be created to maximize person throughput, maximize vehicle throughput (subject to not exceeding the agency's target operating capacity), or to maximize revenues.
Forecast years: See the previous discussion on forecast years for the earlier stages of the project development.
Assumed cumulative projects: The manager would examine the regional long-range transportation plan and select the set of projects that would best meet the needs of the environmental analysis. If input from decisionmakers and stakeholders on these assumptions cannot be obtained early enough in the process, the manager and analyst may try to cover the possibilities through a series of future improvement scenarios. Scenarios cover more bases for the manager, but this advantage comes at the expense of greatly increasing the resources required to complete the analysis.
Study limits: See the previous discussion on study limits for the earlier stages of the project development.
Managing Consistency Between Analysis Tools
The analysis plan for the project clearance stage would incorporate the following techniques to manage consistency across the analysis tools:
Demand model base year and future year network coding for the subarea in the vicinity of the proposed project and its alternatives would be error checked against base year observations and future year plans for network improvements.
Demand model base year volumes would be validated against base year traffic counts. Model parameters and network coding would be modified to improve the validation. The objective would be to make as few changes as possible to the previously calibrated model while achieving the validation targets for the facility and its vicinity.
The demand model-predicted base year performance (average speed) for the project facility would be compared to base year observations and the model adjusted to improve the match.
For the future year no-project baseline forecasts, the predicted average facility performance (average speed) produced by the demand model would be compared to facility average speed predicted by the selected facility operations analysis tool, in this case the HCM-type tool. The demand model speed-flow equation (volume-delay function) would be adjusted for the facility, if necessary, to achieve a more consistent match between the demand model and HCM tool.
Conflicts between simulation results and the HCM tool would be minimized by using simulation to provide supplemental capacity and performance effects for HOT lane access designs not currently incorporated in the HCM tool. The HCM tool inputs (capacity, etc.) for each HOT lane access point would be modified based on the simulation results.
The freeway HCM tool would be used to generate freeway MOEs. The surface street HCM tool would be used to generate surface street MOEs. The demand model would indicate the demands for each HCM model.
The demand model would be used to generate system MOEs for the portion of the entire model network that falls within the significant influence area of the proposed project and its alternatives. The demand model-produced MOEs for the project facility and adjacent streets would be replaced with the HCM tool-predicted project facility MOEs, thus avoiding a potential source of inconsistency (discrepancies between demand model and HCM tool estimates of project facility performance).
During PS&E, construction, and operations, the focus is on the facility. System analyses are no longer required (having been dealt with exhaustively in the project clearance stage). The appropriate analysis tools for these later stages of project development are consequently HCM-type tools, optimization tools, and simulation models.
There is much value in continuing with the tools developed and validated for the project clearance stage. In this case, since an HCM-type tool was selected for project clearance, the same tool would be useful for any lingering design questions not previously addressed.
A simulation model could also be used in the later stages, if it was previously created for the project clearance stage.
At these later stages, consistency is a good quality control check; however, getting the correct answers quickly to design, construction, and operation questions is most important.
In scoping the traffic analysis for the PS&E and construction stages, the manager should recognize that construction detours for overnight facility closures may require significant off-facility analysis of the proposed detour routes. Even if an off-facility HCM-type tool was not required during the project clearance stage, it may be required during the PS&E stage.