The Oil Drum

You’ve hit the problem right between the eyes as with respect to recoverable reserve estimation. I’ve been a career development geologist for 33 years and have spent many thousands of man-days working on the same issues. I suspect you’re aware of many aspects of the process but I’ll expand for everyone’s benefit. There are two basic approaches to recoverable reserve estimation: volumetric and production decline. Volumetric analysis has always (and will continue to be) a rather inaccurate approach. Even when there seems to be enough geologic control to accurately map the volume of oil existing in the reservoir you’re still left with the uncertainly of recovery efficiency. Most large oil reservoir are produced by water-drive: the nature of oil to float on water. As oil is produced the water pushes the remaining oil upward. The problem with analyzing such a dynamic system is the inability to predict with any accuracy how effectively the water pushes the oil out. In reservoirs with very thick oil as little as 15% of the oil is recovered. In light oil reservoirs recoveries can be 60% or higher. But there are more variations possible. The character of the pore connections in one light oil reservoir may allow a 60% recovery while another reservoir with the same quality oil might recover only half that volume. The bottom line: while the in place volume may be fairly accurate, recoverable estimates are much more art then science. But before the reservoir produces this is all you have to work with.

By far the most accurate method to estimate ult rec in such reservoir is by production decline. After X number of years the wells will start producing water with the oil. In most cases, it then becomes relatively easy to estimate ult rec (this is why Saudi has a very good estimate of Ghawar's future production potential even though they won't share that info). There are certain natural power law functions that exert control over the process at this point. But this data isn’t available for years after the discovery is made. From analyzing hundreds of field studies the great majority of ult rec estimate errors were do to optimistic assumptions used in the volumetric analysis. Not fraudulent expectations but utilizing the high side of the different variables. Needless to say, pressure from the management of public companies on the staff to be realistic but estimate high has driven much of the process. Even when outside engineering auditors are used they can still be pushed with technical arguments to be more optimistic (to toot my own horn, I was pretty good at getting the numbers I wanted). And, to be honest, some of those “independent outside auditors” were nothing but “reserve whores” who would gladly skew the numbers if their invoices were high enough. (and, yes, we really would call them that behind their backs).

But I have some questions for Kebab et al. Where is the USGS coming up with their raw data? I have never seen the Survey or any other gov’t agency supplied with any data that would allow any measure of analysis. From a practical matter they couldn’t do such analysis if they did have access to the data. I recall back in the 70’s when the Feds required the Big Oils to supply them with all internal mapping and evaluations of fields in the OCS. The Feds wanted to see if Big Oil was hiding reserves. I knew one of the Survey geologists in charge of collecting the data. He told me that all he and his cohorts could do was log in the data. Analyzing the data with the current staff would take many decades so they didn’t even start the analysis process. I suspect those millions of bit of data are still sitting in a warehouse in New Orleans…if they weren’t destroyed by Katrina.

I’m guessing they are getting reserve estimates from the operators. Same question would go towards your source of data. You guys are very clever and your hard work is greatly appreciated so I hope this next point isn’t taken badly. Are you getting some of you reserve growth numbers from the public companies themselves? And if you are, are you able to distinguish reserve growth via the drill bit vs. growth through acquisitions? Acquisitions have been my focus for almost 20 years. I can only make a wild guess but at least 50% of the reserve growth of the companies I’ve dealt with has come from buying someone else’s field and not discovering new reserves. Thus it wasn’t so much reserve growth (although that’s how they market it to the public) but reserve transfer. I know that sounds so basic so please don’t take offense if you have taken this into account. But that question has been bugging me ever since I started reading about “reserve growth” here and elsewhere.

And, no, Elwood. In 33 years I’ve never seen a company keep two sets of books. I’ve always dealt with public companies and they always threw out the biggest reserve numbers they could w/o getting busted by the SEC (and some actually did get busted). They might secretly talk amongst themselves about how inflated their reserves might be. And don’t even ask me about the scum bag oil promoters I’ve dealt with and the outright lies of exaggerated reserves they would pitch to private investors. For a while I represented such screwed investors. But between the immorality on one side of the table and the ignorance on the other I couldn’t take any more and I no longer play expert witness.

I really don't care for probable, proven, etc. At best, those qualifiers only give you error bars on the estimates. I see all the time where people equate 50% confidence with being 50% off on the estimates. That's like claiming that a 50% chance of rain means that 50% of the ground will be covered by water.

I prefer to use the classical theories of probability using mean values and equating the variance to the square of the mean and letting the model figure out the most probable outcome. See, I used the word "probable" and it has nothing to do with the industry's definition.