I wouldn't worry about the 'project that triggered the NGP2...

I wouldn't worry about the 'project that triggered the NGP2 expansion' - that will never be economic enough to justify the pipeline IMO. At most they might sell a few TJ/d to Barcaldine power station and put a few more down the Cheepie pipeline but it will be a horrendously expensive project - undersaturated, thin coals and with darcy-permeability sandstones in direct contact with them.

If by some miracle they did get it up and running (finding an investor prepared to take the risk of billions of dollars upfront for wells, facilities, water handling and the pipeline) I can't see them ever undercutting CTP.

Back to your model, I'm still not entirely sure where royalties fit into this either - are they included in the $1.50/GJ cost? If so that seems a bit on the cheap side.

Amortisation charges is a great step. I usually use double depreciation for this as it seems to be pretty common in the industry but it doesn't actually make much difference to your final NPV which method you use. It will affect your share price calculation in individual years though. Next step I would recommend is to figure out how they are going to fund it because that will add either interest charges to your NPAT, or dilution to your shares - both of which will affect your PE share price calculation. Probably easiest to assume full debt funding.

Tax is another tricky one. CTP must have built up a decent amount of tax losses over the years but I'll leave that one for the accountants to sort out. Hopefully we can ignore it for the first few years.

All in all, it's a difficult project to model. Any model is only as good as the data that goes into it, for some projects like AWE or COE the data is reasonably easy to pin down but CTP is in a tricky situation where the margin is quite small because of the transport costs. This means the uncertainties in your assumptions (particularly transport cost assumptions) have a large percentage effect on your final net margin, which in turn means a large variation in your calculation of NPV and share price.

This difficulty means I am not losing too much sleep over it - you just have to accept that there is a lot you don't know. I spend quite a bit of time tinkering with my AWE model but I don't worry about it with CTP because I am aware that none of my tinkering really improves its accuracy.

Realistically it probably needs to be modelled probabilistically as a Monte Carlo simulation using @ Risk or Crystal Ball because it's too uncertain for a deterministic model. I should probably do that some day but I suspect it will just give a range of outcomes centered around a mean of 20-30c.