In his commentary defending the pallet concept Van Metre Lund reveals that his Autran system is not intended for crowded urban areas but is more of a suburban connector system. Thus Autran can never be more than a partial solution to our urgent transportation problems. He concedes that the problem of obtaining enough land for extensive access facilities in cities is indeed a serious one. However he doesn’t consider other points of my pallet criticism. Emergency evacuation of guideways, conversion of existing cars (electric heaters would be nice here in Scandinavia), higher energy usage than true dualmode systems and so on.

To strengthen his claim that pallets is a viable concept, Van Metre Lund provides an example of a 10-mile long system, complete with capital cost estimates and everything. While it is perfectly possible that such a system could be financially successful if built anytime soon, I strongly oppose the notion that it could do much to alleviate our upcoming transportation crisis.

In the following will I counter some of Van Metre Lund's claims and expand on his example to clarify why pallet operating systems, such as Autran, are not the best solution to our transport problems, if they are indeed a solution at all. The quoted capital cost estimates are not very specific, so I don’t know if they only cover construction or if land acquisitions are also included. Anyway, I have chosen not to challenge them, but rather use them as they are. It is also unclear what the guideway speed is supposed to be, but the calculations indicate 60 mph, so that is the speed I have used in my calculations.

Van Metre Lund postulates that Autran allows for entry and exit facilities to be located independent of each other. That is of course possible, but only if extra acceleration/deceleration lanes to/from the main guideway are built. And the routing problems are worsened, as it would be necessary to route all pallets from unloading stations onto the main guideway and then off again at a loading station. Not exactly a practical scenario.

Autran supposedly also doesn’t have a pallet storage problem, but has substituted it for a queuing problem, which is exactly the same. Every Autran pallet will need a length of guideway equal to its own length for storage (queuing or parking) when inoperable. Storage can’t be on the main guideway line, nor can it be on acceleration/deceleration lanes, as they have to be ready for traffic all the time. And as only a small percentage of pallets can be in loading positions at any one time, extra lengths of guideway for storage purposes will be necessary. It is true (as noted by J.B.Schneider in my last comment) that they don’t necessarily have to be stored at loading stations, but could be placed in depots along the guideway. However this doesn’t eliminate the problem, only changes it; adding more structures with more acceleration/deceleration lanes, with increased overall system cost as a result.

Returning to the example: I have omitted the intermediate stations, as they don’t add capacity in themselves unless more pallets are also added. They would however make the system more attractive for users and would therefore probably be retained if a system were ever built. For calculations they are only complicating things, so I have omitted them, as well as their costs from the examples. So what we have is a 10-mile long system with 16-lane stations at both ends. In a scenario with 422 pallets as proposed by Van Metre Lund, capital cost would be split with 78% for guideways, 8,5% for pallets and 13,5% for stations. So far so good.

The claim that such a system could launch pallets indefinitely at a 1,2 second rate is however wrong. As the total time to reach the other end station and return with another load would be around 1300 seconds (2x600 seconds transit + 4x20 seconds loading/unloading + 20 seconds lost accelerating/decelerating at both ends) it is easy to deduce that the system could only launch a pallet every 3,1 seconds (1300 seconds/422 pallets) for any sustained period of time. This would produce a line capacity of 1168 pallets per hour per direction (pphpd). Not enough to relieve our congested highways of the strain. Pallets could of course be recycled on a small portion of the guideway (using intermediate stations) to increase line capacity locally, but this would merely reduce line capacity elsewhere. Overall line capacity can’t exceed 1168 pallets per hour per direction (pphpd) in a scenario with 422 pallets.

If we would make an attempt to utilize the 16-lane stations to full capacity, we would need to put more pallets on the guideway. Sixteen loading lanes can launch 48 loaded pallets per minute. Equal to one every 1,25 second. If we therefore issued 1040 pallets (1300 seconds/1,25 second) we could reach a line capacity of 2880 pphpd (3600 seconds/1,25 second). This is the best achievable with 16-lane stations, but still not enough. In this scenario capital cost would be split with 69,5% for guideways, 18,5% for pallets and 12% for stations.

It is possible to match a 3-lane highway, but we would have to double up on pallets and loading lanes: 32-lane end stations and 2080 pallets would produce a theoretical line capacity of 5760 pphpd. This time guideways constitute only 53% of net capital cost, pallets would eat 28,5% while the new, larger stations would require 18,5%. By the way: 2080 pallets (each 20 ft. long) would require 7,88 miles of guideway for storage (queuing or parking) when not needed!

Even a modest city like Copenhagen (1,5 mill. inhabitants) with relatively few congestion problems (so far) have stretches of its highway network approaching 100.000 vehicles per day per direction, so if we want to build a system for the future 5760 pphpd will not be enough. Maybe we will have to redouble the capacity once again: 64-lane end stations and 4160 pallets for a theoretical line capacity of 11.520 pphpd. Providing that one guideway can accommodate so many pallets we would have a capital cost split of only 36% for guideways, 39% for pallets and 25% for our 16-acre end stations.

Considering that a true dualmode system has smaller and simpler access facilities, and doesn’t require pallets, we could build a true dualmode system with similar or larger capacity for approximately half the net capital cost necessary for a pallet operating system. Consider also the problems associated with longer pallet operating systems. If we increase the length of the system in our little example to 20 miles we would need a cool 8000 pallets to maintain a line capacity of 11.520 pphpd. 8000 pallets would require approximately 30 miles of guideway for storage during idle periods (at night).

After all these calculations it is important to remember that the real world is less than perfect, so the real performance of the system described here would be much less impressive. The quoted line capacities can only be reached if patrons would be so kind as to arrive at both end stations at regular intervals in numbers exactly matching line capacity. Unfortunately they won’t. Long waiting times for patrons will be the order of the day, if the system is operating anywhere near its full capacity, and I haven’t even mentioned the wasteful routing of "empties". The pallet distribution problem is not excessive in a small system as described here, but will occur to full effect if a larger network were built. Unfortunately only large networks could make any serious impact on our impending transportation crisis.