Will an airport rail link reduce GHGs?

Given the evident public interest in the idea of a rail link from the  CBD to the airport, I thought I’d look more closely at some of the key rationales for this project, starting with the claim that it would reduce greenhouse gas emissions.

Energy use and emissions (PTUA)

I’ve looked at this issue and, on my admittedly simple calculations, I conclude that the value of greenhouse gas (GHG) savings from a rail line is likely to be minor compared to the probable cost. There are far cheaper ways to offset equivalent emissions than building a rail line.

I looked at this by making the following simplifying assumptions.

First, I assume that a new rail line captures 20% of airport passenger traffic or five million of the current 25 million annual passenger movements at Melbourne Airport. This is double the share captured at either Sydney or Brisbane (around 10%), and almost three times the 7% estimated in feasibility studies.

Second, I assume that all of the current two million passengers using Skybus transfer to the new train (i.e. Skybus ceases to operate) and three million passengers transfer from cars, including taxis.

Third, I assume an average distance of 22 km from the CBD to the airport for bus and train. I assume that the combined average distance travelled to the airport by the cars and taxis that are replaced by train is 35 km.

Fourth, I use estimates of emissions per person kilometre developed by the Public Transport Users Association (here) which were derived from base data prepared by the Australian Greenhouse Office (here). These are based on peak hour conditions and hence assume that bus and train have relatively high passenger loadings and thus low per capita GHG emissions compared to cars (although cars will perform worse due to congestion). The assumed levels of CO2-e per person kilometre are 22g for diesel buses, 14g for electric trains and a whopping 286g for petrol cars.

Fifth, I don’t take account of emissions due to construction of a rail line. It is likely these would completely swamp any savings in operating emissions but I’ll assume that widening of freeways and access roads to accommodate buses in the years ahead has an equivalent impact. I also ignore the fact that Skybus says it currently fully offsets all emissions from its buses

These assumptions are exceedingly generous toward the rail option.

So, multiplying the assumed distances by emissions per passenger kilometre gives emissions per person trip of 0.48 kg for bus, 0.31 kg for train and a massive 10 kg for car. Multiplying these in turn by the assumed number of trips for each mode, shows the carbon dioxide emitted annually by a train carrying 5,000,000 passengers would be 15,400 tonnes. This would replace 9,680 tonnes currently emitted by bus and 300,300 tonnes by car, giving an annual saving of 294,580 tonnes.

Much of the recent discussion around the ETS valued a tonne of carbon at around $20 (here, here, here and here) but I’ll continue being conservative and assume it costs an expensive $50 to offset each tonne of carbon. Thus the annual savings in operating emissions from an airport rail link total $14.7 million.

In other words, the equivalent savings in emissions from a new rail line could be obtained elsewhere for less than $15 million p.a. What would a 22 km rail line cost? A popular estimate seems to be around $400 million if it were connected into the existing inner city rail system but this would probably be slower than Skybus and hence would have little prospect of achieving my assumed patronage level.

The cost of a genuine high speed line could be an order of magnitude higher. For example, according to this report in The Age, the estimated cost of extending the Epping line to South Morang is $562.3 million for a distance of either 3.5 km or 8.5 km, depending on which quote in the story you prefer. Apparently much of the cost is due to the Government’s new policy of not having level crossings. There are a number of major roads, including freeways, that a fast rail line from the airport would have to cross.

Whatever the cost is, it’s probably going to be more than $14.7 million p.a. And let’s not forget I’ve been very generous in my assumptions about rail’s share of traffic.

The relatively low value of savings in GHG emissions is not surprising. In an earlier post I estimated that the value of the savings in GHG from transferring all current Sydney-Melbourne air passengers to a Very Fast Train would only be around $74 million p.a., compared to a capital coast ranging from $27 billion (my estimate) to $40 billion (Greens estimate).

The conclusion’s plain – even on the most generous assumptions, a new rail line is not a cost-effective way of reducing GHG emissions associated with airport surface travel. It needs to be justified on some other basis. In fact a train might very well increase emissions compared to bus – I’ll look at that another time.

Let me emphasise that this is a highly stylised analysis, so some caution is needed in interpreting it. And if someone’s got better numbers or a better methodology, let’s hear it.


7 Comments on “Will an airport rail link reduce GHGs?”

  1. Alan Davies says:

    Note: I’ve corrected an error. The cost of offsetting a tonne of carbon I’ve assumed to be $50 per tonne (not $100/tonne as originally published)

  2. Moss says:

    I agree that the GHG emissions are unlikely to be the main argument for a Airport rail line – especially when you consider the construction aspect (did you know that 7% of global CO2 emissions are from the manufacture of portland cement?).
    One aspect you have not addressed is the air-quality (and it may be difficult to do this). Obviously removing cars from the road would shift pollution to the electricity generation point.

    • Alan Davies says:

      Re your point about CO2 emissions during construction, here’s a relevant quote from an article in The Journal of Urban Economics re energy:

      “In addition, the construction and expansion of new and existing rail systems is very energy intensive. For instance, Tri-County Metropolitan Transit Agency claims that under the best case scenario, the proposed north light-rail line in Portland, Oregon would save the equivalent of 7875 gallons of gasoline per day. But the agency also calculates the energy cost of building the line to be 32 million gallons of gasoline. Thus,even using the most optimistic estimates and assuming no depreciation of the capital stock, it would take a minimum of 15 years to even begin to achieve energy savings—and concomitant reductions in emissions—from this rail line”.

      Winston C and Maheshri V, On the Social Desirability of Urban Rail Transit Systems, Journal of Urban Economics, 62 (2007) 362-382

  3. […] I’ve pointed out before, the savings in carbon emissions from substituting a train for Skybus could be bought for vastly […]

  4. […] previously discussed the issue of a Melbourne Airport-CBD rail link here, here, here, here, here and here). Possibly related posts: (automatically generated)More on Melbourne […]

  5. […] time hasn’t come. Not yet. I’ve outlined the case against an airport rail link before (here, here, here, here, here, here and here), but in summary the key objections […]

  6. […] Ninth, even if a train were to win an unprecedented share of all airport ground travel (say 20% – which would put it way ahead of Brisbane and Sydney), this would be an extraordinarily expensive way of reducing GHG emissions. […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s