Managing Transport Demand To Boost Capacity

Most of us would like to be able to travel when, where and how we want to and for the transport system to be managed in such a way that there will always be enough capacity to allow us all these choices. The problem with this  “capacity management” approach is that a lot of money would have to be spent providing capacity that is only used for a very limited time of the day.  Without this extra spending we still have to continue putting up with congested roads and overloaded public transport during peak hours.

Required capacity could be reduced by managing the “when”, “how” and “where” choices. This post looks at some  “demand management”strategies that might be used to reduce peak capacity requirements  These strategies offer rapid, low cost  ways of getting more from the transport infrastructure we already have. It was concluded that a rapid, low cost doubling of capacity is not an impossible dream.

DETAILS:

Australian mega-cities are all suffering from transport system capacity problems. For too many people, peak hour travel involves very slow driving along congested roads or standing up for long periods in overloaded public transport.  Worse still, the overloaded system contributes to long delays if there is an accident or other disruptions to the transport system.

In the past we have tended to respond to increases in peak demand by increasing system capacity to meet this demand.  This approach may be OK while there are low cost options available to meet new requirements.   It is no longer OK if we have got to a point where further increases in capacity will require things like bulldozing buildings to widen roads, adding extra train tracks, digging tunnels and other expensive, brute force options.

Even the old “divert more people to public transport” can reach a point where increasing public transport capacity becomes a lot more expensive than just buying a few more trains and buses. For example, in Sydney, some experts are suggesting that “Sydney trains could be fixed by halving car registration and cutting tolls.”

There is a limit to the capacity of train lines and stations. Building extra train lines and higher capacity stations is never going to be cheap, particularly if the infrastructure can only be installed underground or up in the air. CBD bus stations don’t come cheap either.

A key part of the problem is that both fixed and mobile transport infrastructure capacity requirements depends on peak demand, not average demand. Fig 1 is a graph of trips vs time of day is based on travel to and from the Brisbane CBD. Similar patterns would apply in other large cities.

Fig 1: Trips vs Time of Day – Brisbane CBD

The graph highlights the extent to which the narrowness of the morning peak contributes to the capacity needed to allow us to travel when we want to. This post looks at two specific strategies for reducing peak demand:“flattening the peaks” and “reducing daily commutes.” In both cases it is assumed, for convenience, that the distribution of travel time and commute method would stay the same. (This does not mean that changing these distributions should not be part of a broader plan for improving transport capacity.)

FLATTENING THE PEAKS

If you look at fig 1 morning public transport flow into the CBD the peak occurs at about 8 AM with about 50,000 trips into the CBD happening at peak time. In theory, this peak could be reduced by 25% by changing the time when trips occur so that, at no time, the number of trips exceeds 37,500. Some rough estimates based on fig 1 suggest that, to meet this limit, some, but not all public transport commuters who currently travel earlier than the peak would have travel 6 mins earlier than they do now. Those who currently travel later than the peak may also have to travel 6 mins later than they do now.

To halve the peak to 25,000 trips the changes to travel time would have to be 18 mins earlier or later.

Similar rough estimates could be made from fig 1 on what changes in travel time would have to be made to reduce peak car travel into the CBD. For the morning car curve a 25% reduction in peak would need start time to be moved by about 9 mins.

It is assumed that time spent at work will not be affected much by the above changes. This means that afternoon peaks will also be reduced if the morning peak is reduced. (They are already lower than the morning peaks.)

NOTE: It is emphasized that the above figures are based on a particular set of figures and may be different for different for other dates and locations. In addition, on some routes, what may need to be controlled is arrivals at some congestion point some way out from the CBD rather than arrivals at the CBD.

How could these changes be achieved?

One of the simpler ways of flattening the peaks would be to change workplace start time distribution to the extent that this is practical. Changing start times does not require significant expenditure into complex controls and could be initiated by convincing some, but not all, employers to change their start time policies.

It should be kept in mind during discussions that there will be different limits on what changes individual employers and employees are willing or able to accept:

1. The nature of some businesses means that they will need to continue starting at a specific time. (Think a morning coffee shop.)
2. Some businesses will be able to change their start times but will want all their employees to change by the same amount. (Think of many factories.)
3. Some businesses will want members of an individual team to start at the same time but would have little problem with different teams deciding to start at different times.
4. Some businesses could accept considerable variation in the start times of many of their employees as long as they put in the hours required to get the job done.
5. Some employees would be able to do part of their work before or after travelling to/from the workplace. (Think: Planning the day, writing reports, checking emails and making phone calls)
6. Some employees will have limits on what changes would be acceptable for personal reasons. (Think: Can’t leave home before they get the kids off to school.)
7. Some people will live where they can get to work without contributing to congestion or public transport overload. (Think: People living within walking or e-scooter riding distance from work or those who use routes that are not congested.)
8. Some people would choose to travel outside of peak hours if they were able to choose flexi-time.

In some cases it may be practical for individual employers to control individual start times to match the overall start time distribution target. In other cases, achieving the overall distribution target will require co-operation between employers to get the overall mix right.

The above figures suggest that, right now, we could get significant improvements without anyone being forced to change starting time by more than 15 minutes. 

The system may have to become more sophisticated and authoritarian as pressure on transport capacity rises. Even then, the cost of these changes should be negligible and there will be cost savings due to reduced public transport fleet size, better public transport fleet utilization and the deferral of expenditure on transport infrastructure.

It should be remembered that what is important is the managing  of peak hour commutes along routes that suffer from congestion or overloaded public transport systems rather that when people actually start work.   

How hard to get started?

There is no reason why individual employers and employees could not start doing something now to start improving transport capacity.  Steps might include:

1. Government decides transport demand management is worth pursueing and:
• Decides what work centers and overloaded routes and services to start with.
• Collects demand data and travel time data vs time of day for these routes.
• Converts this data into forms that will allow commuters to see whether moving their travel time.
• Consult employers and unions about potential attractions and problems. 
2. Employers identify employees who:
• Have jobs that would not be affected by changes in start times.
• Travel along routes that are affected by peak hour overload.
• Who have no strong personal reasons for not changing start times.
3. Employers identify current practice that might need to change to make changing start times more attractive.  (Ex: A "Meetings must start  and finish between..." rule is needed so that early starters are not kept at work after their normal leave time and late starters miss out on the start to a meeting.)
4. Start voluntarily moving start times for employees and employers who want to give it a go.

REDUCING COMMUTES 

Reducing weekly commutes and/or average commute distances will help reduce transport emissions and other environmental damage.

Part of the problem with transport in mega cities is that most city workers belong to a group that all work the same old 5 day Monday to Friday commute roster.  Both the rosters described below address this problem

5 Day, Monday to Friday COMMUTE Roster
This approach would suit people who need to work Monday to Friday AND do not need to commute 5 days per week because they can do part of their work at or near to home.  (Most people whose job involves frequent dealing with other people” could do this from home using phones, emails and Skype equivalents.)

A 5 day commute roster might involve:

1. Workforce is split into A and B commute crews.
2. Crew A may commute during peak travel times Mon, Wed and Fri.
3. Crew B may commute during peak travel times Tues and Thurs.
4. Workers who are not rostered for the days peak commute would either work at home and/or commute outside of peak commute times if they really do need to come into work on that day.

In theory the crews might swap rosters every second week.  However, for people dealing with outsiders there are attractions in having the same commute roster each week to avoid confusion.

Moving someone to a 5 day commute roster that averages 2.5 commutes per week would halve their weekly peak hour commutes, commute emissions, commute travel distance and travel time.  In addition, employers may save money because less work space would be required. 

NOTE:  5 day commute rosters don't have to average 2.5 commutes per week.  Any roster that averages more than zero days per week working at or close to home will produce some benefits.

How Hard to Get Started?

Commute rosters are very easy to set up because an employer could start a trial with only one employee working one day per week and then adding extra employees and increased number of days later on.
Governments could help by:

1. Campaigns to encourage employers to try some working at home as part of the work mix.
2. Information that will help working at home boost both employee and employer satisfaction with the experience.
3. Monitor the distribution of working at home days across the week with the aim of maximizing the contribution this distribution makes to transport.  (A day with very few people working at home will have higher peak loading.

7 Day WORK Rosters

7 day work rosters provide a way of reducing the daily commutes of people whose job can only be done at the workplace.  People who are part of one of these rosters will be split into at least two crews with half the workers on the roster being at work on any one day. 

 Moving someone to a 7 day roster would halve their Monday to Friday peak hour commutes and reduce commute emissions and travel distance by 30%.  It may also save employers money because less space and equipment may be required because the number of people at the workplace on any one day might be reduced.  In addition, changing to a 7 day roster could boost production by 40% if the workers kept to the same working hours per day or double if the workers worked the same weekly hours.

There is nothing radical about this 7 day rosters proposal. Many industries work 7 days per week and use multiple crews to cover the 7 days. Some of these rosters are toxic to the workers and their families while others are quite attractive. For example, a number of mines that I worked at in Central Qld had each crew working 4 days on 4 days off. (12 hr shifts.) This was a very popular roster. The mines that had this particular roster had very stable workforces with good morale. The roster gave the workers more days off per week than Mon to Fri while avoiding the long periods away from home that were a feature of some of the more toxic rosters.  

Weekend days off is an issue when comparing different 7 day rosters. A 4 on 4 off roster gives, over a 7 week cycle, two full weekends off plus two other weekends where the crew members had either Sat or Sun off. Other rosters could give different patterns of weekend time off. For example a 7 on, 7 off roster could be arranged so that the crews always had Sat or Sun off every weekend.

Simply splitting the original single crew into A and B crews does have a number of disadvantages:

1. There is no overlap when both crews are at work (Bad for work communication.)
2. The only time the crews will have time off together is when they are on holidays or on strike. (Not good if family members are in different crews and not good for communities.)

The above problems with two crew systems can be reduced by setting up a 4 crew system where:

1. A and B crew alternate.
2. C and D crew alternate with C crew starting at least one day after A. This overlapping means that there is at least indirect communication between all crews and that a person working A or B crew will also share time off with C and D. (Gives a better chance of family members sharing time off and better communication within the community.)

The family and community problems will also be reduced if different employers use different rosters or simply start roster cycles on different days. This will increase the number of people who share some rostered days off. 

To reduce confusion it is important that, for all employers “A” crew describes crews that are rostered on and off on the same days.

Small businesses and teams may decide being part of a roster is attractive while choosing to all work in the same crew.

My personal preference would be 4 on 4 off with 4 crews and with C crew lagging 2 days behind A crew. Other workers may have different priorities.  We can play for hours developing rosters that have particular attractions.

It would not take long to move at least some people and employers to 7 day rosters. Costs and benefits would depend on what, if any, increases in wages are involved and the potential savings and increased sales for employers.

How Hard to Get Started?

There are already a wide range of businesses that have people working at weekends.  Some of these businesses will use casual workers to cover the weekend, some will use overtime and some will already use rosters of some sort.  

Getting more people and businesses to switch from working Monday to Friday to a 7 day roster could be challenging.  Some people would find the rosters described above attractive because of longer breaks and time off when there are fewer people at beaches etc.  Others may be put off by needing to work longer days to maintain their weekly working hours and/or because it may reduce the time they have with some friends and family.

Moving to 7 day rosters may be attractive to businesses that need to boost production.  However, in some cases the nature of the business may mean that moving to a 7 day roster will require employing a second crew.  (In other cases the man-hours per unit of production will remain the same.  In this case the number of workers at work on any one day would actually drop and the number of employees will only rise if production is increased.)
There may be some cases where it is possible for a small number of people who like the idea of working a 7 day roster to switch.  Think about people who need to be at work to do their job, can do their job alone or with a very small number of other people and who don't need to work all the time with people working Monday to Friday.   The move to 7 day rosters is most likely to happen where a business needs extra production or 7 day per week coverage and/or there are workers who like the idea and can move to a 7 day roster without reducing efficiency.
Governments can help by talking about the potential benefits of 7 day rosters to transport and quality of life and publishing interviews of people who are already on 7 day rosters 

CONCLUSION:

Australian mega cities have already reached a point where the transport system cannot cope with the expectation that the system must have the capacity for people to travel when they want to, how they want to and where they want to.   

This post demonstrates that some forms of transport demand management have the potential to dramatically increase the capacity of the existing system without spending a lot of money or having to wait for years for any significant benefits to be felt.  In some cases there will be genuine reasons why specific employers or employees should not be expected to do some of the things suggested but the transport crisis has reached a point where there must be a very good reason why start times should not be adjusted to reduce peak transport demand or serious consideration be given to other strategies that will reduce peak demand.

Perhaps employers should be asked what they are going to do to reduce the number of their employees and contractors who contribute to peak demand.