Scenarios, ranges, the theoretical and empirical: Why the difference matters, but isn’t a stop-light.

There is one theme that I constantly fall back on, in examining many things, particularly where innovation is involved.? The importance in any scale-up - of an empirical foundation. Likely that's by virtue of geology - my profession - being what it is. An essentially "fuzzy" science (some would say warm and cuddly). Fuzzy in the sense of not being the result of putting seven numbers in a formula, and a unique result coming out of the black box. It has no monopoly on that, but it is a particularly "fuzzy" near end-member.

Moreover, fuzzy in the sense of being the result of thousands of variables, some more important than others, that conspire together. The art of geology and the science of geology is very much one of discerning which of those thousands of variables matter most to what we want to do with it, in any given place, at any given time. And how that interacts with everything else we have to do topside to "harvest" any reward, and please those we want to deliver something to.

In looking at such things, we can never be very certain about much in the sense of precise singular values, so we use probabilistic distributions to help capture what might "be". That's a fancy way of saying ranges of inputs, that are weighted in some way, shape or form, based on theory and/or experience.

The problem is that each attempt to do so - whether we realise it or not - hangs off a series of assumptions, a model, a framework - what we term a "scenario". Sometimes we don’t even realise we are applying single scenarios that might have other possibilities.? Each of these scenarios envisage and apply certain base assumptions around what our inputs will be. They may hang off theoretical laws of chemistry and physics (which we don't want to play around with too much if we want to stay credible). They may hang off assumptions around geometries of things underground. They may hang off prices and politics. The may hang off how time influences things.

After all, we are not just interested in what we can do. We are interested in what we can do that delivers a price competitive product to a customer for a certain use over the project lifetime, in the timeframe they want it. In a way that is better than what constantly evolving competitions can do and/or could do. We can never know all that, so we make assumptions, and that's OK. We have no alternative path to making assumptions. If we decide not to make assumptions, we do precisely nothing and that helps no-one. BUT we do have to appreciate how strewn with assumptions things are.

Geology is a fantastic science, and one of the reasons it is greatly appreciated across so many disciplines, is that its students appreciate better how great nature, and time, and people's interaction with both, is at throwing our "scenario" assumptions out the window. Dispassionately. We are very sophisticated at modelling particular scenarios, but it is when unsuspected scenarios kick-in that we are less good. Many studies in many geoscience subjects and projects have shown this repeatedly. ?

The geologist then, always recognises a need to think about many scenarios – the “out of the box” things that might throw us a wide-ball. Not just to think about them, but to estimate how probable they are relative to other scenarios.? We do our best, but we are never perfect at that.? Sometimes things come at us either underground, or "topside", that just blow our expectations out of the water.? Sometimes those are technical, sometimes those are “other”.?

Who knew for example that an earthquake in Japan would almost overnight be the straw that broke the camel’s back for German nuclear.? Who knew that the discovery of newts would stop Boris Johnson building a new swimming pool at his Oxfordshire Manor.? We might say, but that’s not geology, but we are not just doing geology.? As we said, we are delivering to a customer’s expectation.?

Similarly, small quirks of fate that happen can shift public opinions almost overnight too, not necessarily logically.?? Noting, and saying it is illogical doesn’t make impact any less real or go away.? Public opinion is not any less real for sometimes being flawed - even if much of the time it gets a lot right.

Some things technically are just hard to know.? We can predict what many technical things going wrong might look like, modelling what we believe are “worst case scenarios” but what the cost of remediation will be in ten years time, and what the customer’s attitude to it will be is harder.? And all the while they will be looking at the competition to see what alternatives are possible and how they are evolving, and not just whether they are cheaper, but whether they are quicker and/or less hassle.?? Sometimes there is a quite significant premium a customer will pay if they can sort it out once for a long, long, time and more or less sit back and forget about it.? Options have to compete on that basis as well as deliverables, cost and speed to deployment. ?Geologically involved options are not always in that "sit back and relax" category.?? Sometimes the rewards can genuinely be worth the hassles – sometimes emphatically so, but sometimes it takes the empirical data to truly know.?

What all this means is not that we stop modelling, or stop making assumptions, or stop planning on what we know so far based on theory and the experience we have had. We don’t throw up our hands and say it is too hard, because experience tells us there is value to be had in persevering and experimenting.? It does however mean we recognise an important difference between theoretical and empirical. The theoretical uses what we think we understand, applies it to certain base scenario assumptions, and makes logical projections. We can use ranges and make multiple scenarios and blend their outcomes, trying to the best of our abilities to capture a good sampling of the most likely situations.? All well and good.??We can however, never know in advance that we have captured all reasonable, possible, scenarios.?

The empirical in contrast doesn't care about theory. It records what actually happened. There are no assumptions involved.?? It is what everything that is, does to influence what we are trying to do, irrespective of whether we imagined it might happen that way or not.? ?

Doing something new always starts at the theoretical-end with less empirical information. That's OK. It’s no show-stopper to doing new things.? BUT we just need to recognise it, recognise that a maturity isn't quite there yet, which things with a lot of empirical data do have. When something has hundreds of empirical data points taken from full-life-cycle repeated projects in many different places, it puts them in a different league of confidence to new things that haven't got there yet.

The latter, innovations, are still worth doing - we want to advance and not stay stuck in the old if the new might ultimately be way better. We just need to acknowledge the difference between theoretical and empirical and hence how much we should rely on things accordingly. And we need lots of empirical to truly discern reliability. If we do the innovative and theoretical, fantastic, but to do it with eyes wide open, and to be hesitant, careful, in populating a total replacement of everything we know with the new – until it has a stronger empirical footing. Not to deny entertaining the possibility, but to wait until it is confirmed empirically before we swop out the foundations.

?