This paper describes the bottom-up by-field global oil forecast model built and maintained by the author over the last 15 years, originally within BP, and subsequently independently.
The model is based on a database of oil production from the world’s individual oil fields since 1992, and projects production to 2040. The forecast is for mobile and synthetic crude, including Canadian tar sands and Orinoco oil, as well as ‘light-tight’ oil, but excluding condensate and NGLs, and oil from kerogen, coal, gas or biomass. Initial production data came from oil industry sources, but are now supplemented by data from those OECD countries that supply reliable field data, company reports, and the field data in the Oil and Gas Journal (‘O&GJ’) database. In many cases reported field production data are aggregated (for example, by operator), or are missing (in which case interpolation or extrapolation are used); and values for small fields sometimes have to be calculated by subtraction from country totals.
Oil reserves data are drawn from a variety of sources, and only proved-plus-probable (‘2P’) data are used where available. O&GJ reserves data are not used, as these are 1P data.
The modelling generates production profiles by field, where these reflect growth, plateau and decline phases; and where if the actual profiles are not known, they are generated by analogue to neighbouring fields. In addition, a field’s total production to 2040 is checked against the reserves data where available.
Estimates of yet-to-find oil are based on a number of sources, including past USGS estimates, but modified by subsequent discovery volumes, actual discovery rates and professional opinion. The resulting estimates are fed into a simple model of discovery rate.
The model pays particular attention to undeveloped (‘fallow’) fields, possible technical improvements, and production of the various ‘non-conventional’ oils.
The model forecasts that global oil production (as defined above) can potentially climb fairly rapidly (above the likely rate of demand increase) to reach a peak round 2023, at a little under 110 Mb/d. In practice production will rise to meet demand, and hence peak somewhat later.
The paper concludes with discussion of uncertainty in the data (particularly of production from currently fallow fields); historical experience of the model’s accuracy; the need to add economic considerations to the model; and the strengths and weaknesses of such bottom-up ‘by-field’ models.
Miller, R. G. (2015) A Bottom-up Model of Future Global Oil Supply. The Oil Age 1 (2) 39-56.
Download PDF – FreeCLICK HERE