Reserve Growth for dummies

To explain how reserve growth works with oil, let's start by forgetting about oil and looking at beans.

Imagine that you've just inherited a house from your gran and you find a cupboard loaded with tins of baked beans. There are four shelves, each with three layers of tins, 10 tins wide and 20 tins deep. You don't want to take all of the tins out to count them so you make an estimate of how many there are; in fact, you make three estimates.

First you count all the tins you can actually see and you decide that is your minimum estimate – the least amount of tins that there must be in the cupboard. Let's say you counted 300. You call that P95 because there is a 95% chance that there are at least 300 tins of beans. (Why not 100%? Well, some of the tins might be empty but you won't know until you open them.)

Then you work out the maximum estimate. If every layer that you can't see fully was full, there would be 2,050 tins in total so you call this P5. (Why only 5%? Since you can't see the whole of the cupboard, there's a possibility that there are some more tins squeezed into the back somewhere.)

Finally you make an estimate of what it is likely to be the total, the P50. This is the value that is just as likely to be more as less. You look at the gaps in the shelves, think how your gran would have stacked them and, let's say, go for 1,700. Being omniscient, I happen to know that is almost right – the 'ultimate' for the Bean field is 1,800.

So we have a P95 of 300, a P50 of 1700 and a P5 of 2050. Notice that the P50 is not simply halfway between the two values.

So you start living in the house and eating the beans. Every week, you eat 10 tins ('production') and that means every week the total cache of beans (your 'reserves') goes down by 10. After the first week, there are 1,790 left (although you don't know that). After a year, there are 1280. Your R/P ratio is 1800/520 = just under three and a half years.

So where does reserve growth come in? If you used the P50 to estimate your total tins, there wouldn't be too much of a problem. Your first estimate was only 100 out. As you eat the beans and reveal more of the shelves, you could make a better estimate of what was originally there but it wouldn't change dramatically. As you near the end, you would eventually know exactly how much the original cache was but your P50 would only move from 1700 to 1800, a slight increase in the reserves. (Because the P50 is just as likely to be an overestimate as an underestimate, it might have gone from 1900 to 1800. The change would still have been small, less that 10%.

But what happens if you used the P95 as your working estimate? Initially the value is 300 but, since there are really 1800 tins in there, your P95 - your estimate of what is almost certain to be there - increases by large amounts over time since there is a big difference between 300 and 1800. Let's say after the first month, you change your P95 to 400 since you can now see more of the originally hidden tins. What it looks like is that you have found 90 new tins that month (100 extra tins minus the 10 you used) since your estimate is increasing whereas, in reality, your cache has decreased by 10.

What you really should do is backdate the P95 to the day you discovered the cupboard so it is not the reserves that are growing but the original cache. Otherwise you will think you are getting more beans when, in reality, you are losing them. This is what ASPO do to get their estimates of oil reserves and why they differ from the official figures such as BP's.

This applies particularly to oil since it is the P95 that is often used in estimating reserves. The US SEC (Securities and Exchange Commission) insist on using P10, and the oil companies prefer to use the lower estimates since it means that they can always report increasing reserves to their shareholders. The result is that we get a false impression of how much oil there is in the world and how bleak the future actually looks.

While real oil reserves are more complicated than tins of bean - they take into account the costs of extracting the oil as well as what is actually there - you should at least now understand what reserve growth is and how careful we should be about attaching too much importance to it.