# How to Break a North Sea Project Cost Estimator

How do you know what a North Sea development project will cost? Well, most of us have our rules of thumb, some a tad more sophisticated than others. They range from just knowing that North Sea capex per barrel has been running at nearly $30/bbl for the last few years to sophisticated computer models based on construction rates and equipment weights.

I am going to show you one model I made this weekend to cross check our own capital cost estimates and then I will show you how to break that model and get a cost per barrel that makes a North Sea project worth investing in. For, if the North Sea is ever going to regain its competitiveness, we need to invest. As Trevor Garlick, BP's head honcho in the North Sea, said when BP launched its project to revitalise the ETAP fields last week "** The solution to the North Sea is to keep investing. If we stop investing we’re finished.**"

Trevor is right, we need to be renewing the North Sea not just patching up what we have. But when the oil price is less than $50/bbl spending $30/bbl upfront to bring on new production doesn't make sense. That's why everyone who looks at oil and gas investment opportunities says "*Your project is very interesting, but n**o, the North Sea isn't for us*". They don't believe an estimate of $17.50/bbl capex, their internal project cost estimator is saying "It will really be $30/bbl" and the project won't work. That's because most finance people's cost estimation model is not much more sophisticated than an average cost per barrel.

So what I am going to show you is a brand new UKCS standalone project cost estimator that works with just five variables. The astute among you will recall that John von Neumann famously said "*With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.*"

Well it turns out there is a least squares Fourier sine series with five constants that does a pretty good job of fitting an elephant, and with eleven constants you can hear her trumpet across the Savannah. Nip over to Wolfram Mathematica to make your own elephant, you will need to download a plug-in but its worth it to have fun with Fourier sine series.

As it happens, with five variables you can do a pretty good job of matching North Sea projects. The five variables I am using are the reserves, the peak off take rate, the water depth, the reservoir depth and the number of wells, oh and I keep track of whether the wells are subsea or on a platform. A little bit of research can find those variables for pretty much any project. To make my predictor I have just used a multivariate regression model to match some recent projects, here is how my predicted capex matches actual capex for that set of projects.

The R squared for that prediction is something like 99%, but that isn't so hard to achieve with five variables and only eight observations, see von Neumann's elephant above. Let's unpack the analysis a bit and see if it is doing any better than a simple average North Sea capex per barrel. Here is the same data on a per barrel basis; what this chart is showing is how incorporating the reservoir depth, the water depth, the off take rate and the number of wells is helping us predict the capex per bbl. The R squared for this is 82%; based on that I would say that the predictor is probably quite a helpful tool, but we should be cautious especially if we use it with values that are bigger or smaller than the range of observations.

Let's see how good a job it does on some upcoming projects, here is the first chart with Pilot (our project) Bentley, Rosebank and Lancaster added on. If you want, for a small fee I can let you know where your project sits, for a large fee I can tell you how to break the prediction model and cut the development costs of your project substantially.

Pilot is the red triangle just to the left of the $3 billion line, we currently think we can get the project done for c. $2.5 billion, but we included in that estimate a 17% reduction in upstream capital costs in line with what WoodMac and IHS have seen in early 2015. If we hadn't done that we would have been bang on the line. So this whole exercise has been worthwhile, our baseline capex per barrel is pretty much what you would expect given the shallow water, shallow reservoir, offtake rate and number of wells.

But as I said at the start, the whole point of working on a project development plan is to try and break the prediction, to do better than business as usual. We did it in the North Sea before in the early nineties when we started developing smaller fields like Harding and Andrew. They seem like pretty chunky oilfields now, but back then all the cost curves told us that the capex per barrel had to go up as the oilfield size came down. Well, in the end the Harding development plan came up with a cost per barrel that matched (on an inflation adjusted basis) what the Forties development cost per barrel had been. It broke the cost curves of the day.

It was applying horizontal drilling technology and working in collaboration with contractors and suppliers that broke the old models in the early nineties; this time round we need both to rediscover those old approaches and to find new ways of breaking today's capex model.

I promised at the beginning of this article I would show you how to do that and the truth is there isn't just one way, there is a multitude of ways, but they all involve being open to new technologies and new approaches. You can't break the model if you do the same things we have always done and hope that the cost will be lower, that's just magical thinking and it won't work.

__conductor sharing wellhead__from Cameron. For our project we need lots of wells, and on average every well slot costs a lot of money – about half of what it costs to actually drill the well itself. If we can use one slot for a couple of producers, or maybe even three injectors then we can eliminate over half our well slots and make a real difference to our costs.

I can already hear the questions and the concerns, but if it can be made to work on wells being drilled on the Magnus field, then for our shallow wells it should be possible too. To fit in two producers we would probably need 46" conductors (that is because our producers need a slightly larger casing size to accommodate hydraulic submersible pumps) but with 46" conductors we could fit in three steam injection wells so that is even more efficient. We could go from needing 42 slots to less than 20, that would probably take another $250 million or 10% off our capital cost estimate.

There are lots of great ideas and innovations out there, some of them even work. The art of breaking the North Sea capex model is choosing the technologies that will work, and that will enable breakthroughs in the cost of development.