Is there a case for an overnight sleeper train between Auckland and Wellington?

There has been a bit of discussion recently about the potential for a night train between Auckland and Wellington, so I thought I would look into what a North Island sleeper service might involve and whether it could be a good idea.

I consider myself something of a first-hand specialist on sleeper trains. I love long distance train travel and in my backpacking days I, er, slept my way right across Australia, Asia, Europe and America. I’ve done everything from six hours ‘sleeping’ in a chair on the night train to the salt flats in Bolivia, to an almighty six-day train trip from Mongolia to Moscow on the famed Trans-Siberian railroad.

For the passenger, the ability to sleep through long overnight travel in comfort is the main attraction of a sleeper train. This is obviously a good alternative to a long overnight car or bus trip, which can range from unpleasant on a bus, to downright dangerous behind the wheel. But it can also be a good alternative to a much shorter daytime flight: once you factor in eight hours or so spent sleeping anyway and the time taken getting to and from remote airports and checking in, a night train can be effectively faster than flying on the right corridor.

Who takes sleeper trains?

So what is the point of a night train and who are the target markets? Based on these experiences travelling overseas I think there are five kinds of people who use night trains:

1. People with little money who want to get the best deal they can, students and backpackers, including travelling overnight sleeping in a chair. They are looking to travel cheaper than flying or driving and are happy to waste their time and comfort to do so.
2. ?Enthusiasts who do it for the romance and experience of train travel, most of whom are happy to spend far more time and money than flying, but expect comfortable/luxurious private cabins and premium food and drink.
3. Mid-range tourists and business users who travel overnight to save the cost of a nights accommodation and to maximise their daylight hours travelling. They expect reasonable comfort and access to reasonable food and drink and facilities like showers and luggage.
4. People who are socially/environmentally conscious and try to avoid flying or driving too much. They have similar expectations to the mid range tourists.
5. People who are afraid of flying (I swear half of Amtrak users are in this group, the fear of flying is strong in the American Midwest!)

Looking at this, there’s probably a reasonable demand base from a broad range of users, which is a good thing for a business if you can serve them all with the same product.

Caledonian sleeper double room with ensuite. Seat61.com

So which kind of route works?

Any good long-distance trains need to be anchored at both ends with major demand centres that drive two-way traffic across the year. This means trains between big cities. Smaller towns and tourist destinations that are seasonal or have little non-tourist activity outside of a couple of peak weeks will struggle to support trains alone, so these are best as stops on the way between to strong anchors. The best route for a sleeper train is between a pair of main cities that have a few other towns on the way.

What sort of schedule should sleeper trains have?

In my experience night trains work best on routes that are ten to twelve hours long, with a departure around 7pm to 9pm and an arrival around 7am to 9am. Less than that and it is too rushed and compressed, you don’t have time to sleep for eight hours, and you end up either leaving after midnight or arriving at 4am, etc. But go much longer than twelve hours and you feel you are wasting time by leaving too early or arriving too late, and feel the need for lounge cars and other things to occupy your waking time. Interestingly there are quite a few examples of night trains that actually stop and park on a siding for several hours to extend a route that is too short to give a convenient timetable.

A key success factor for sleeper trains, in my experience, is that the timetables need to have a long non-stop section in the middle during the overnight hours. ?This is because stopping and starting trains and blowing whistles at stations constantly wakes everyone up every half an hour all night, and very few people want to get on and off a train between about 10pm and 6am anyway. Having caught the odd one at 3am I can say there isn’t much sleeping involved that night, and much more the next. Also, running every night of the week, and in both directions every night, is ideal as it gives people the opportunity to travel when they need to. The fact is most business travelers, weekenders, tourists on short trips, students etc. don’t actually have the ability to wait around a day or two for when the train is actually running. The only people who can do that are premium tourists who book well in advance and plan their whole trip around the train.

A scheduled train both ways between Auckland and Wellington every night?

Accordingly there is at least one route in New Zealand that clearly meets the criteria for a successful night train, Auckland - Wellington. There are some other routes that might work but I’ll stick to this most obvious one for now.

This has the strongest demand for travel between two main centres in the country, linking the largest city in the country to the third largest, and the economic capital to the political and cultural capital. It is also on a route that can take in two main centres along the way, Hamilton and Palmerston North. This also has the right distance and route length at around eleven hours travel time each way. This gives it the right timing to create an overnight route that’s long enough for sleep, but its still short enough that it can leave and arrive at convenient hours.

The Auckland to Wellington day train route, rome2rio.com

My suggestion is to run two trains a night, one each way. And to run both directions every night of the week. The southbound train would depart Auckland at 8pm, picking up extra passengers at Pukekohe, Huntly and Hamilton. After the final pickup stop at Hamilton at 10:30pm it would the run non-stop overnight to Palmerston North, with the first arrival at 6:30am, followed by stop at Paraparaumu and arriving in Wellington at 8:15am. In the northbound direction it would do the same thing in reverse, Depart Wellington at 8pm, stop at Paraparaumu then a last pick up at Palmerston North at 9:45pm, before running non stop overnight to Hamilton for the first drop off at 5:45am, followed by Huntly and Pukekohe for an arrival in Auckland at 8:15am. These non-stop sections in between allow passengers the chance to sleep as the train makes its way through the thinly populated centre of the North Island. This does mean the train skips potential stops such as Te Kuiti, Taumaranui and Ohakune, but on a night train these would be reached deep in the middle of the night so it’s unlikely many people would want to get on or off anyway.

If you look closely, you can see I’ve timetabled Hamilton to Palmerston north at eight hours, which is actually about two hours more than it takes on a direct run. This is so the trains can park in a siding for a couple of hours early in the morning (Taumaranui or National Park village would be the place to do that). This has a couple of reasons: it stretches the run time so that the departure at one end isn’t too late or the arrivals at the other aren’t too early, it gives passengers a full eight hours in the middle without others getting on and off in which to sleep, it gives a good place for the train crews to take a break and swap staff, and it builds some fat into the timetable so that if there are any delays getting away they can make up the time overnight.

Service classes of cabins and seats

My suggestion is an Auckland-Wellington sleeper with a three-class service with three price brackets, to sever a range of users markets. In my experience a three-class offering in the most common around the world. So these three price ranges would be:

• First class sleeper, with luxury private cabins with one or two people per cabin and ensuite bathroom. This would come with a sit-down restaurant-style food and drink service, and personal host service. This targets the luxury tourists and some of the business travellers.
• Standard class sleeper, with more compact mid-range cabins with up to four people per cabin, (e.g. bunks with two up two down), or maybe compact two-bunk and single options. These would probably have shared toilet and shower facilities, and self-service café counter food and drink options. This is for the mid-range tourists and business users and the environmentally conscious types.
• Economy class: open carriage with large reclining seats in 2+1 arrangement (NZ trains are quite narrow so getting four across would be a squeeze). These would likewise have shared toilet and self-service café food and drink options. This is for the backpackers, class trips and budget conscious.?

Train carriages and capacity

Based on similar sized trains overseas each train carriage would fit about eight cabins, or 54 seats in the economy carriage. So first class is up to 16 people per carriage, standard class is 32 per carriage, and economy is 54 seated. I’ll assume each train consists of seven carriages and a locomotive, to easily fit in existing station platforms. I’ve not assumed there is need for a separate baggage car as people will take luggage with them into the cabin or above their seat. This gives us something like this for each train, with capacity for 220 people if completely full:

• 1x first class cabin carriage (16 passengers each),
• 3x standard class cabin carriages (96 passengers)
• 2x economy seated carriage (108 passengers),
• 1x café counter/restaurant car, with generator.
• plus a locomotive.

What would the ticket prices be?

In terms of ticket prices, based on international experience the standard class should be priced similar to a standard full price airfare on the same route, in this case about $200 one way for a standard class berth in a cabin. Per person, first class should be a bit more than double standard class, at about $450 a head (or $900 per cabin), while economy should be about half the cost of standard class at around $99 each way for a recliner. If you add up these fares levels and the number of berths and seats proposed, each one-way train run could bring in up to $40,000 of revenue if fully booked out.

What about the cost of buying and running these trains?

Luckily New Zealand has the capacity to make these trains. We have a large supply of spare carriages from the old Auckland fleet, and the industry within the country to design, modify and refit them for other purposes. Kiwirail has done this several times for its acclaimed tourist train carriages, and most recently for the Te Huia commuter train from Hamilton to Auckland. There are some parts that come from abroad and other considerations, but they could be designed and built within 12 months of getting sign off. ?Indeed from the business case for Te Huia?we have a very good idea of what it would cost to set up the trainsets for an Auckland-Wellington sleeper train.

Based on these recent train developments by Kiwirail, I expect we can budget $1.2m per carriage for the cabins and café cars, $1.0m for the seating cars and $2.5m per locomotive refit. This means each seven-carriage trainset would cost $10.5m including the locomotive. With two trainsets required to run both ways that’s a capital cost of $21m to build the fleet. ?If we assume these new trains can be stored, maintained and serviced at the existing Kiwirail facilities in Auckland and Wellington, the fleet should be the only major capital cost to set up the service.

Carriage being rebuilt at the Hutt Workshops for Te Huia, Kiwirail

In terms of operating cost, again the Te Huia business case gives us a great indicator. Kiwirail is charging $5.0m per year to fuel, staff, manage, operate and maintain the Te Huia trains. This buys twenty-two one-way trips a week on the Hamilton Auckland run. Doing the math, we can see this equates to 1,716 train-hours of operation per year for five million dollars, or a cost of $2,900 per service-hour per train. These costs are very high by international standards, but that’s another story. For now our night train concept would have ten and a bit hours of service time per train (plus the layover), which amounts to a little over $29,000 per train, per night.

?With two trains each night, running every night of the year, the operating cost of the night train service would be $21.6m each year.?

To recap, the bottom line

By international standards Auckland to Wellington is a good candidate for a sleeper trains, with the right sort of route length, timing and demand drivers to be a success. In this post I have proposed a schedule for a nightly sleeper train both ways between Auckland-Hamilton and Palmerston North-Wellington, starting each end around 8pm and arriving at the final terminus around 8am the next day, with an eight-hour non-stop period through the night. The two trains would each have capacity for 220 passengers in a mix of premium and standard cabins and seats, across a seven-carriage locomotive hauled train. The trains could be re-built from carriages and locomotives already in New Zealand, and be operating within about 12 months.

With these trainsets and schedule a North Island night train would have the potential to replace up to 150,000 long distance car trips or flights per year, and in the order of 75 million vehicle-kilometres-travelled.

The capital cost to set up the trains would be around $21m dollars up front, with ongoing costs of $22m per year to run. With fares roughly equivalent to flying between Auckland and Wellington, it would need to achieve average occupancy of 74% full each night across the year to break even. While maintaining very high occupancy levels every night of the year is perhaps unrealistic, these quick sums do suggest there is at least the possibility of a decent business case, especially when non-fiscal factors like emissions, travel time savings and business productivity are also considered.

So, what say you? Is this idea worth studying further?