Over the course of well over two hundred hours with the game, I have run into situations where I felt like I wasn't getting the performance I wanted from freight elevators. The statistics of the freight elevators can be looked up within the game, but I thought I'd create an experiment to test them.

The objective would be to create a somewhat realistic configuration of a freight elevator, loaders, conveyors and storage so I could actually measure the throughput of the whole system.

First, some ground rules. Since we are trying to measure the maximum throughput of a system, I decided to use the best components, so I have a Freight Elevator IV and Conveyor IV components. The loaders and storage have no upgrades.

Next ground rule. When using a freight elevator there is always an initial start up delay as the internal belt goes from the starting level to the final level. That delay disappears as soon as there is a continuous supply of items to be lifted/lowered and so is not relevant to my test. To minimize the effect of that delay I placed the two levels of the elevator close to one another.

Here is my initial configuration. There is a storage at each end and I'm using a pair of loaders at each level, placed close together to minimize the impact of the travel time of the item on the conveyor.


Initial configuration

The entire setup is powered by one transformer. I'm using the start button on the transformer to power up the loaders which begins the transfer of items.

For the items I just happened to have some Cleaning Bots nearby. I put 320 of them into the lower storage and the test would measure how quickly they were moved to the upper storage.


Cleaning Bots, my items

The test consisted of four phases. The first would have only a single pair of loaders at each level. The second would have two pairs each, the third three and the final would have four pairs at each level.


The final configuration

At the start of each phase I would have the 320 Cleaning Bots in the lower storage. I started my video recording and enabled the transformer which powered up the loaders and started the transfer. I moved the camera from the transformer to the upper storage and stopped the recording after all 320 Cleaning Bots reached the upper storage. I then watched the video, pausing and recording the time from when the transformer was powered up till the last of the Cleaning Bots reached the upper storage.

After watching the videos I came up with the following results.

Phase one, 18 seconds
Phase two, 9 seconds
Phase three, 9 seconds
Phase four, 9 seconds

So having a third or fourth loader pair at each level had no effect. This is not unexpected, once the stats of the components are noted. Below is a list of the throughput of each component:

Conveyor IV, 1280/min or 21.33/sec
Loader, 1800/min or 30/sec
Freight Elevator IV, 2560/min or 42.66/sec

While one might note that the results of the experiment seem obvious, it is always useful to periodically run tests to confirm that there aren't any unexpected or hidden factors influencing the performance of a system.

In this case the throughput of a loader isn't a limiting factor in moving items on a freight elevator. Loaders are faster than Conveyor IV belts (currently the fastest in the game) and Freight Elevator IV elevators (also currently the best). It is the latter two items that limit the throughput.

In practical terms, if you are using class IV freight elevators and have the latest class IV conveyors, you need to have two loaders at each of the start and final levels to get the maximum throughput. Adding more loaders and belts does not increase the throughput.

If you are not using the latest class IV components, then a review of the stats of the component would likely give you the best configuration. However, a quick review of the different classes of components shows that if the class of the freight elevator matches the class of the conveyor, then you will always need two loaders with conveyors at each level of the elevator. A good rule of thumb to keep in mind.