The Bicycle Utility Ratio:

A Suitability Metric Based on Cost/Benefit Analysis

for Utilitarian Cyclists

Steven G. Goodridge

North Carolina Coalition for Bicycle Driving

4/6/2002

Introduction

A recent trend in bicycle planning has involved the creation of maps showing the degree of "suitability" of different roads for cycling. The "suitability" metrics such as "Bicycle Level of Service" [1] and "Bicycle Compatibility Index"[2, 3] typically used for such maps are based mostly on the comfort preferences of casual recreational cyclists, and to a lesser extent safety. These metrics fail to consider the utilitarian value of the roadway for cyclists who depend on bicycling to reach important destinations. Since recreational cyclists like to stay away from traffic, but important destinations always attract traffic, these metrics skew the results of the mapping process to discourage utilitarian cyclists – especially those who do not have cars – from using useful roads to go to popular places. In order to serve the needs and rights of utilitarian cyclists, consideration of the suitability of a roadway for bicycle transportation should be based on a cost/benefit analysis.

A new suitability metric called the Bicycle Utility Ratio, or BUR, is developed below to incorporate both costs and benefits to citizens who use bicycles for basic transportation. The BUR is defined as the relative usefulness of the roadway for bicycle access to destinations divided by the relative hazards of traveling on the roadway by bicycle:

**BUR = (Usefulness)/(Hazards) ** [1]

Usefulness

The relative usefulness of a roadway depends on the number and importance of the destinations it serves as well as the efficiency of access it provides. Straight, direct routes with priority over cross traffic are more useful than circuitous routes with frequent stops. At first it may seem difficult to estimate the relative usefulness of one roadway compared to another; however, people vote for useful roadways by adjusting their travel habits. The best estimate of the usefulness of a roadway for the population at large is the average daily traffic on the road, or ADT. (For the simplicity of the following model, pedestrian volumes will be ignored but do amount to significant transportation demand in some places.)

We could simply use the total vehicular ADT to determine the usefulness of a road for cycling, but some roads feature a large proportion of long-distance motorist through traffic. These longer distance trips are a result of the convenience of motoring and do not translate well to the potential number of bicycle trips. In order to filter out the effects of long-distance motoring trips on the utility of a road for cycling, the ADT could be weighted based on the percentage of trips that access nearby destinations within easy bicycling range. The average utilitarian bicycle-commuting trip in the United States is about three miles, although many bicycle trips are longer. If a detailed traffic analysis of the roadway and surrounding areas is available, a weighting function Kbu can be developed to properly weight the ADT. For example, Kbu may incorporate 100% of the trips that access trip endpoints within three miles of the roadway, 25% of the trips that begin or end within five miles, and 5% of trips that begin or end within ten miles. The resulting usefulness function is

**Usefulness = Kbu*ADT** [2]

Traffic Hazards

Bicycle transportation is not a particularly dangerous activity when compared to other physical activities and other travel modes. However, the safety of cycling is commonly described as a concern for transportation planners and cyclists. The vast majority of the popular concern is usually centered on the potential hazard caused by motor vehicles. This risk is affected by the behavior of road users, the design of the roadway, and the volume of motor traffic. For this model let’s assume the behavior of users is not a variable. The risk from motor vehicles may be modeled as the product of the traffic hazard per motor vehicle (Kth) on a particular road due to the roadway design, and the number of vehicles, ADT. If we only consider motor vehicle traffic hazards, then

**Hazards = Kth*ADT** [3]

Incorporating Equation [2] and Equation [3] into [1] we have

**BUR = (Kbu*ADT)/(Kth*ADT) = Kbu/Kth **[4]

We see that since the ADT cancels out, this model for the BUR depends entirely upon the design of the roadway and the percentage of short versus long-distance traffic on the road, which is in turn affected by road design and local land use patterns.

Surface/Fall Hazards

Studies of real-world bicycle transportation risks have shown that the vast majority of injuries to cyclists operating on roadways involve falls and other mishaps not involving moving motor vehicles. If we incorporate a factor Kfh to represent the fall hazard for a particular road design and surface condition, the BUR model equation becomes

**BUR = (Kbu*ADT)/(Kfh + Kth*ADT) **[5]

Or,

**BUR = (Kbu)/(Kfh/ADT + Kth) **[6]

For high ADT volumes this model of the BUR approaches a constant value equal to the value in Equation [4]. For lower traffic volumes, the BUR *decreases*. This is the opposite of the result for the traditional BLOS and BCI metrics, but makes complete sense to utilitarian cyclists (especially those who do not have cars) because low-traffic-volume roads serve fewer useful destinations yet still involve the risk of falls, the time and effort of pedaling, and exposure to the elements. Thus the best ways to increase the BUR for any street involve increasing the percentage of near-endpoint traffic using it (Kbu) and improving the design of the street to reduce fall hazards (Kfh) and the traffic hazard per vehicle (Kth). Short-cut streets connecting residential and commercial land uses have very high Kbu. Increased street redundancy increases the Kbu of minor through streets while reducing the Kbu of major arterials. Well-maintained streets and streets with wide outside lanes have low Kfh values. Streets with raised center medians, left turn lanes, and wide outside lanes have reduced Kth.

Example:

An existing street has an ADT of 8000 vehicles per day, half of which are long-distance through trips and another half of which involve nearby trip endpoints. After a new shopping center, apartment complex, professional park, and residential subdivision are added around an activity center location, the ADT increases to 12,000 vehicles per day. What happens to the Bicycle Utility Ratio?

Let’s assume that somehow we know values for Kbu = 0.5, and Kfh and Kth. Let’s also assume that the total fall hazard, in terms of disability-adjusted years of life and medical expenses, is estimated to be twice the traffic hazard at the original traffic volume of 8000 vehicles per day.

Thus we have, using Equation 5:

**BUR1 = (Kbu*ADT)/(Kfh + Kth*ADT) **

= (0.5*8000)/(2* Kth*8000 + Kth*8000)

= 0.5/(3*Kth)

= 0.167/Kth.

Since all of the 4000 trips induced by the development on the street are endpoint-related, Kbu is increased to 8000/12000 or 0.67. Let’s also assume that Kfh and Kth have not changed. Now we have

BUR2 = (Kbu*ADT)/(Kfh + Kth*ADT)

= (0.67*12000)/(2*Kth*8000 + Kth*12000)

= 0.67(12000)/(28000*Kth)

= 0.287/Kth.

Here we see that the added activity center development has increased the Bicycle Utility Ratio. This matches observed trends in utilitarian bicycle transportation: that when an area urbanizes, the number of utilitarian bicycle trips increases both in absolute numbers and as a percentage of the total traffic volume.

Limitations of the Model

This model does not fully capture the preferences of avid cyclists who like to travel long distances on arterial roads for commuting, although the motor vehicle ADT may correlate with this. It also does not incorporate the effects of combined recreational/utilitarian bicycle trips by people who own cars and switch to automobile travel when they find elevated traffic volumes to be unpleasant when traveling by bicycle. The model does not capture the effect that congestion-reduced motor vehicle speeds have on increased bicycle transportation. Lastly, it does not capture the nonlinear effects of traffic volume on traffic hazards; for example, on narrow two-lane roads the hazards of unsafe passing increase in relation to the product of the volume of overtaking traffic and the volume of oncoming traffic.

Conclusions

The Bicycle Utility Ratio incorporates the popularity of trip endpoints into the suitability measurement of a street as used by regular utilitarian cyclists. The BUR model reflects the trend of increased bicycle transportation that results as an area urbanizes and average trip distances on streets become shorter. The model shows the effects of local development and relative availability of access routes on bicycle transportation demand. Given that increased traffic volume correlates with increased bicycle transportation by dedicated utilitarian cyclists, analysis of the BUR model suggests that cyclists should be accommodated on high-volume roads that provide access to important destinations. This accommodation should be designed to maximize cyclists’ efficiency while reducing falls and car-bike collisions through proper roadway engineering, education and law enforcement. In general, conventional suitability mapping projects have little value to utilitarian cyclists except where they identify useful shortcuts and pleasant but efficient alternate routes that may not be obvious on ordinary maps.

[1] Landis, Bruce, "Real-Time Human Perceptions: Toward a Bicycle Level of Service*,*"* Transportation Research Record* 1578 (Washington DC, Transportation Research Board, 1997).

[2] *Development of the Bicycle Compatibility Index: A Level of Service Concept*, *Final Report*, FHWA-RD-98-072 (1998).

[3] *Development of the Bicycle Compatibility Index: A Level of Service Concept, Implementation Manual*, FHWA-RD-98-095 (1998).