HERS-ST Discussion

Rather than making use of the HPMS Submittal data for the needs analysis, OrDOT has routinely built HPMS formatted datasets from scratch, making full use of the enormous transportation system database that exists for the state. The dataset is "100% sample", representing the entire length of each state highway on the roadway system. All data fields are tracked by highway number and beginning milepoint location in an effort to insure that each record represents a homogeneous section of highway. Several specific fields are further identified as control fields, such that whenever data in any one of these fields change then a new record is created. As a result the record length is not standard. The control fields are as follows:

County Code,
Urbanized Code,
Functional Class,
Facility Type,
AADT,
Number of Lanes,
Median Type,
Median Width,
Lane Width,
Shoulder Width,
Shoulder Type,
Speed Limit,
K-Factor,
Terrain Type,
Pavement Condition (Tie to PMS),
Bridge Location (Tie to BMS),
Region Boundary, or,
District Boundary.

This data development process creates large datasets. The dataset used for the recent highway plan contained 24,386 records. However, with the new computing power available to the user, the number of records is not the disadvantage that it was years ago, when the analyst started the run before going home for the evening and evaluated the analysis in the morning when they returned to work the following day.

Since the needs models do not require all HPMS data elements to be inputted, "dummy" values are insert into the unused fields. The subsequent analysis is not project level, but the overall quality of the analysis is superior to that of simply using the submittal dataset, which relies on statistical expansion factors to enhance analysis.

However, the data development process leads to one of the primary problems that we have in the overall analysis process. This deals with what I call the "Sphere of Influence" that is associated with any traffic control device (i.e., signal or stop sign) on the roadway system. In evaluating capacity and subsequent congestion on a system at or near some type of traffic control device, there is a distance out from that particular device in which traffic flow is directly effected due to the device. Because of the queuing effect of the device (and etc), free flow speeds are not attained.

The record length various in the data development process such that one or more records can be created within each traffic control device's "sphere of influence." Additionally, since each record is analyzed individually, the influence of the traffic control device is not propagated ahead (or back) to the next (or previous) record(s) that may still be effected by the traffic control device. Meaning that both the congestion and capacity for the neighboring records will most likely have a higher capacity and lower congestion levels attributed to them, since the analysis does not know a device exists, than would otherwise normally exist. There is no way to correctly capture and portray the true capacity and congestion on the system, or to identify the true improvements.

OrDOT has a similar problem in several situations where off ramps for the interstate becomes congested during certain times of the day and that the congestion is propagated back onto the travel lanes. Under this circumstance, drivers are decelerating on the freeway before they ever get to the off ramp, rather than on the off ramp itself. Subsequently, there is congestion on the mainline that is directly associated with the ramps, but there doesn't seem to be a way to capture that congestion for the needs analysis.

The point of discussion here is to think about and suggest workarounds.

Ø County Code, Ø Urbanized Code, Ø Functional Class, Ø Facility Type, Ø AADT, Ø Number of Lanes

• County Code,
• Urbanized Code,
• Functional Class,
• Facility Type,
• AADT,
• Number of Lanes, Ø Median Type, Ø Median Width, Ø Lane Width, Ø Shoulder Width