The calculations needed to find out the date of the world conventional oil and gas production peaks, and anticipated rates at which non-conventional oil and gas can compensate, are fairly straightforward. Here they are summarised, but for more details see the many papers by Campbell and Laherrère in the open literature, or the original consultancy reports listed in this web site under Further Information>Consultancy Reports .
We must start with some definitions.
‘Hydrocarbons’ refers to oil or gas. Coal, though it contains some hydrogen, is not usually considered a hydrocarbon.
Conventional oil is defined here (and fairly generally) as oil produced by primary or secondary recovery methods (specifically: own pressure, physical lift, water flood, and water or natural gas pressure maintenance.). However, this definition is not universal. On this definition, conventional oil currently accounts for about 95% of all oil production, with some 1-2% coming from enhanced recovery, and a further 2-3% from heavy oils, and tar sands. (In addition, an additional 10% of the 'liquids' supply is provided by natural gas liquids from gas fields.)
There is no agreed terminology for ‘reserves’, ‘resources’ etc., but here we use the fairly common definitions of:
- Resource: All of the mineral, whether discovered or not, whether recoverable or not.
- Recoverable resource: That part of the resource that is recoverable under certain assumptions (usually not stated) on price and technology level.
- Reserves: That part of the recoverable resource that has been located, but not yet used.
- Yet-to-Find: That part of the recoverable resource that has not yet been located.
- Ultimately recoverable reserves (the ‘ultimate’): The original endowment of reserves, hence this is the same as the recoverable resource.
Ultimate = Cumulative production + Reserves + Yet-to-Find.
Reserves are generally classified as proved, probable, or possible; where these, usually, are seen as additive, so the largest amount of reserves judged reasonably likely are the (proved + probable + possible) reserves. Alternatively, one can quote reserves as 95% likely (P95); 50% likely (P50) or 5% likely (P5). Here we use P50 reserves and (proved & probable) reserves as synonymous; this correspondence may not be strictly correct, but given the uncertainty in real-world reserves quantities (see text), appears justified. (See the text, also, for the extraordinary unreliability of published ‘proved’ reserves.)
Units used are:
Mb: million barrels; Mb/d: million barrels per day.
Gb: giga (billion) barrels.
Tcf: trillion cubic feet. (top)
The analysis carried out by Colin Campbell and Jean Laherrère in the 1995 study for Petroconsultants (now IHS Energy Petroconsultants) on conventional oil was as follows:
(a). Estimation of ‘P50’ oil reserves, by country. (‘P50’ reserves are those with a notional 50% probability, i.e., being equally likely to see downward revision as upward revision with time.) These estimates were generated by taking the reserves data from the Petroconsultants’ data base, but adjusting:
- in the light of the authors’ extensive geological knowledge;
- on the basis of a variety of reasonableness tests. A key one of these is to plot a field’s production vs. its cumulative production. For most fields, once in decline, this plot gives a good check of the field’s likely ultimately recoverable reserves. (For fields in the former Soviet Union (FSU), for example, this approach shows that the reserves of many fields are significantly over-reported.)
(b). Generation of estimates of oil yet-to-find. This analysis was on a basin-by-basin basis, where appropriate; and mostly used a range of statistical approaches, essentially based on the discovery data to date, to estimate the quantities of conventional oil likely be found within a reasonable exploration time-frame (for example, assuming twice as many wildcats as already drilled in a basin).
(c). Addition of cumulative production, P50 reserves, and to yet-to-find, to give an estimate of each country’s ‘ultimate’ (i.e., ultimately recoverable reserves).
(d). Modelling each country’s future production by:
- if already past peak, by declining production at the existing decline rate (i.e., by a fixed percentage of the remaining recoverable resource);
- if prior to peak, by increasing production at an annual growth rate until cumulative production equals half that country’s ultimate, and thereafter declining production at the then-existing decline rate;
- in the case of the Middle-East ‘swing’ producers, calculating their production, subject to their own resource limits, using a small number of ‘geo-political’ scenarios. (top)
There are very large amounts of non-conventional oil and gas in the world, so the issue here is not the size of the resources base, but the rate and cost at which these hydrocarbons can be produced.
The Resource Base of all Hydrocarbons
- The blocks in this Figure are all to-scale. Data are given in billion of barrels of oil (or oil’s energy equivalent in the case of gas), Gboe.
- Reserves are industry data, i.e. (proved & probable) reserves.
- Abbreviations: CONV: Conventional
- The Figure shows the resources in-place, and the proportion thought to be recoverable under current and medium-term technology.
- For conventional oil and conventional gas, the hatched bars show the amount consumed to-date.
- Shows (by dotted lines, and smaller-font underlined figures in italics), for conventional oil and gas, and enhanced recovery and non-conventional oil, the quantities that will be consumed and found over the next 10 years, at the present consumption and discovery rates.
- Note: Recoverable gas hydrate quantities may be large, but probably are not; see e.g., papers by Laherrère.
- Sources: Based on data in F. Harper, and Perrodon at al., (see Further Information>Consultancy Reports).
The calculations on the rate that non-conventional oil can expand are based on known projects, and on reasonable extrapolations into the future.
As mentioned in Summary, the various types of non-conventional oil face a number of fundamental constraints (including cost, investment requirements, energy requirement, energy payback times, other input requirements, and pollutants, including elevated CO2 levels) that limit the rate that they can be brought on-stream.
More work is needed to define the various growth rates that these non-conventional oils might achieve. (top)
The production profile of conventional gas is less certain than that for conventional oil, but will be similar to that shown, and is resource-limited in the medium term. Here it is modelled by XXXXXXX. (top)
For the future production of non-conventional gas, the same remarks apply as for non-conventional oil. (top)
Periodically, the United States Geological Survey (USGS) makes an estimate of the amounts of the global oil and gas that are yet-to-find, and adds these to past production and reserves data from IHS Energy Petroconsultants, to arrive at estimates for the world’s original endowments of conventional oil and conventional gas by basin (and also aggregated by country). The last such survey was in 2000, and is available free of charge.
There are a number of important comments to make about these data (see, for example, the ODAC publication in Energy Policy, February 2002), but provided the data are handled with caution, they can be used to calculate the conventional oil peak dates by country.
The way to do this is as follows:
(i). For all countries that are clearly past peak, determine the total quantity of conventional oil in each case that was consumed at the peak.
(ii). Compare these figures with the USGS 2000 estimates of each country’s ‘ultimate’ (i.e., original endowment of conventional oil), and list each country’s ‘peak percentage’ (i.e., quantity of oil used at peak divided by the USGS figure for that country’s ‘ultimate’, expressed as a percentage.)
(iii). Drop from this list those countries (or adjust their percentages) where you judge special situations make this percentage unreliable.
(iv). Determine a representative percentage across the countries included.
(v).. Apply this percentage to the remaining countries’ ultimates (i.e. those not yet clearly past peak) to determine their likely peak dates.
The assumption behind this method is that the USGS survey approach, while finding significantly higher ultimates (at ‘mean’ probability) than estimates based on past discovery rate, is reasonably consistent between countries in its estimating approach. (top)