That's a great question.
Let's start with time, time for, in our case, a train to go from an origin to a destination and to return. We measure that in hours. The quantity is moot.
There is time over a particular subdivision. When we talk about granularity, to us that means not measuring things in terms of the system-wide averages. In our particular case, from southeast B.C. to Vancouver, that's not that useful to us. In fact, it's not useful at all. We would want to know, for our specific good, the time, the quantity, the availability of locomotives, the availability of cars. Gosh, why locomotives? What is the redundant capacity? We don't plan to 100% perfection, so what's available to us, should we need it? What's the contingent capacity on cars? Are there some available? Are they all being utilized? We go into a great deal of detail on that in our submission. There is time across a particular subdivision.
When we talk about the issue of congestion for someone like Teck, we could aggregate it, but we want to know what's moving in the corridor in which our goods are travelling. You could aggregate the rest of the traffic. You could do it by car type. You could do it by length of train, and so on. But when our particular good now merges with the other goods, how are they behaving in conjunction with each other? We do that. How it's happening in January might be different from how it happens in August. There's a seasonality to it too.
There's labour capacity. How many additional crews do you have? You have to measure, on a very granular basis, the supply chain in order to understand whether you have adequate capacity—that's the denominator—and what's actually moving—that's the numerator.