USGS Assessment 2000

ABSTRACT: This is a review of some of strengths and weaknesses of the USGS 2000 Report. Briefly put: Despite its shortcomings (discussed later), IMO, it is an innovative and potentially very useful document.


I. EIA AEO98 Report: The EIA is the public relations department of the USDOE and, IMO, it's "Annual Energy Outlook 1998" (AEO98) is, at best, not a useful document. Per Tom and Jay, "Reverse-engineering" is an accurate description. No further comment is needed.

II. USGS 2000 Report: A colleague and I recently spent several weeks reviewing the "USGS World Petroleum Assessment 2000 -- Description and Results" comprising some 32,000 pages and available free (1) on 4 CDs, and (2) on the USGS web site. Fortunately, most of the important information is contained in about 300 pages. Briefly said: We find USGS 2000 report to be a landmark assessment of the world "Grown Conventional Petroleum Endowment". Moreover, it contains an innovative new and evolving (but yet unproven!) method for predicting the 'growth' of undiscovered petroleum fields and reserves. That said, the USGS 2000 report can be very useful, but it must be used with care and caution!

This review describes and gives examples of some of the strengths and weaknesses of the USGS 2000 report. Don't worry about understanding ALL about the report. Nobody does. In fact, it was done by a large team of geologists and supported by as many statisticians and technicians. Moreover, all the basic petroleum data they used, in their own words, "is proprietary". i.e. We can't even review it.

Definitions: "Grown Conventional Petroleum Endowment (GCPE): The sum of known petroleum volume (cumulative production plus remaining reserves), the mean of the undiscovered volume, and additions to reserves by reserve growth." Also, "BBO" and "Gb" mean the same thing.

The "world petroleum reserves" and the "world undiscovered petroleum volume" are both "grown" by a "30-Year Growth Multiplier" that is statistically derived from (sic) the US petroleum production and reserves historical data. This is both a major strength and THE MAJOR WEAKNESS of the USGS 2000 report, as discussed below.


The USGS 2000 divides the world into 8 geographic regions:

  1. Former Soviet Union
  2. Middle East and North Africa
  3. Asia Pacific
  4. Europe
  5. North America
  6. Central and South America
  7. Sub-Saharan Africa and Antarctica
  8. South Asia

Each region is divided into 'Provinces', Provinces are divided into 'Total Petroleum Systems' (TPS), and the TPS are divided into 'Assessment Units' (AU). The study began in 1995 and was released in 2000 -- hence "USGS 2000". Several databases were used. The USGS 2000 world assessment did NOT study the United States petroleum system, but it DID include previous assessments of the US by a USGS report (1995) and a Minerals Management System report (1996). All USGS production and reserves data are normalized to 1/1/96. Petroleum geologists (mostly from the USGS, but a few contractors as well) assessed each AU for the 30-year period from 1995 to 2025. Then, as one USGS geologist later revealed, "We tossed the assessments over the fence to the statisticians."

NB: Contributors to USGS 2000 include several respected petroleum geologists whose names frequently appear (or are mentioned) on this energy-resources newsgroup. Nonetheless, CARE AND CAUTION is advised in using the USGS Assessment 2000.


The USGS 2000 divides the petroleum assessments into 'categories of probability': F95, F50 (i.e. median), F5, and Mean (i.e. arithmetic mean). "F" means fractile, as defined by the USGS:

"Probability: Probability (including both geologic and accessibility probabilities) of at least one field equal to or greater than the minimum assessed field size. Results are fully risked estimates. ... F95 represents a 95 percent chance of at least the amount tabulated. Other fractiles are are defined similarly."

EXAMPLE #1: The Executive Summary (ES), Table 1, "World Level Summary [including the US] of Petroleum Estimates for Undiscovered Conventional Petroleum and Reserve Growth for Oil, Gas, and Natural Gas Liquids (NGL)," lists the following world total "Grown Conventional Oil Endowment (GCOE)" in billion barrels (Gb):

TOTAL GCOE at F95 = (approx.) 2,000 Gb

TOTAL GCOE at F50 = (approx.) 2,700 Gb

TOTAL GCOE at F5 = (approx.) 4,900 Gb

TOTAL GCOE Mean = (approx.) 3,000 Gb

The values of the world GCOE (above) range from 2,000 Gb for F95 to 4,900 Gb for F5. Moreover, because the world oil production in 1995 was 24.7 Gb and the cumulative production at end-1995 was 539 Gb, this suggested to many, IMO, that that we had a fat cushion somewhere between 1,461 Gb and 4,361 Gb for future production, i.e. GCOE/Prod = 59 to 177 years before the world would "run out of oil". (Kind of like the ubiquitous, but misleading, P/R ratio.) If true, then there was/is plenty of time for the scientists and engineers to discover and tap vast amounts of positive 'net energy' (EROEI) from e.g. 'oil shale', methane hydrates, deuterium-helium.3 fusion, and the quantum vacuum. Thus the USGS 2000 report, IMO, is likely to mislead all those people who don't dig deeper.


BACKGROUND: Four previous USGS assessments were made, but each contained only a SINGLE CATEGORY for "Undiscovered Conventional Oil Resources". USGS 2000 Table RV-1, "Summary of Previous USGS World Petroleum Assessments (which include the United States)" gives the following previous USGS assessments of the mean "Total Oil Resources". The "Effective Date of Assessment" is shown first followed by "Total Oil Resources" in billions of barrels (BBO):

1/1/81: 1,719 BBO.
1/1/85: 1,744 BBO.
1/1/90: 2,171 BBO.
1/1/93: 2,273 BBO.

But then the SINGLE CATEGORY of "Undiscovered Conventional Resources" (see above) was expanded the USGS 2000 report into TWO CATEGORIES, namely "Undiscovered Conventional" AND "Reserve Growth (Conventional)" as given in Table 1 of the Executive Summary (chapter ES). As a result, the USGS 1/1/93 value of 2,273 BBO (listed above) "grew" to a whopping (sic) 3,021 BBO in the USGS 2000 report.

So how did this happen? The key concepts are "growth", "grown" and "adjusted upward". Three definitions will be useful (there are many more in the USGS 2000 Glossary, chapter GL).

"Grown Conventional Petroleum Endowment (GCPE): This was previously defined.

"Grown Petroleum Volume: Known petroleum adjusted upward to account for future reserve growth. For this assessment, 30 years of reserve growth is considered."

"Known Petroleum Volume: The sum of cumulative production and remaining reserves as reported in the databases used in this assessment."


The USGS report dated 1/1/93 (see Table RV-1, USGS 2000) lists world mean "Undiscovered Conventional Oil Resources" at 471 BBO (i.e. ONE CATEGORY ONLY). In contrast, the USGS 2000 (Table 1, Executive Summary, ES) lists world mean "Undiscovered" at 1,251 BBO (sic) comprising the following TWO CATEGORIES: (1) "Undiscovered Conventional" = 649 BBO, and (2) "Reserve growth (conventional)" = 612 BBO, wherein 649 + 612 = 1,261 BBO. An increase (i.e. "growth") of (sic) 780 BBO between the USGS 1/1/93 report and the USGS 2000 report (i.e. 780 = 1,251 - 471).

In this example #2 I'll only only show how the "Reserve Growth (conventional)" of 612 BBO (above) was calculated in the USGS 2000 report for the world (excluding the US) by the following formula:

Reserve Growth (conventional) = (Known Petroleum Volume) * (30-Year Growth Multiplier) = ( 859 + 539 ) * (0.44) = 612 BBO (neglect rounding)

The Known Petroleum Volume (above) comes from Table 1, chapter ES. The 30-Year Growth Multiplier (above) comes from Table RGApp-1 and Figure RG-1.

One BIG "elephant in the ointment" is that the 30-Year Growth Multipliers for world oil (and gas and NGL) are all derived from the US Lower-48 petroleum production data. Then the Growth Multipliers are applied in one-final-fell-swoop to the WHOLE WORLD, as shown in the above example. (Shades of Zapp [or was it McKelvey? ca. 1955] who extrapolated the Texas oil production data to the entire US onshore Lower-48 and came up with an EUR of about 600 BBO -- wherein it now appears to be only about 200 BBO.)

The USGS is well aware of this, and other, shortcomings with this initial attempt to forecast world petroleum growth --

"The forecast of world potential reserve growth described here is considered to be preliminary. Much work remains to be done on the subject of world potential reserve growth. The present study is an attempt to provide a numerical hypothesis for world potential reserve growth that is valuable in itself, and will perhaps act as a stimulus for discussion and research aimed at reducing the uncertainty of world reserve-growth estimates." (USGS 2000, RG-4)

BOTTOM LINE: The USGS 2000 report, IMO, is an innovative and valuable contribution to world petroleum forecasting. As the methodology for assessing and "growing fields and reserves" evolves, the uncertainty will be reduced, and the estimates for the world Grown Conventional Oil Endowment will converge somewhere between 2000 and 2200 BBO (i.e. near the F95 estimate in the USGS 2000 report). The peak of world oil production is within sight.


  1. As far as I know, Colin J Campbell months ago was first to observe that the USGS 2000 F95 values for the volume of world conventional oil are realistic.
  2. More recently, Jay Hanson suggested the importance of taking a close look at North American natural gas (NG) supplies. This has now been done and will be reported in a separate email.

Rich Duncan, 2-20-01