Reserves uncertainty: some historical trends and wider implications

It is a well-known fact that the petroleum industry generally does a poor job in estimating hydrocarbon reserves. Judging from the comparison of pre-drill estimates and post-drill outcomes, reserves from exploration discoveries in USA, for example, tend to be grossly overestimated (P.R. Rose, pers. comm.; Capen 1992). A similar message comes from the Shell Group (Sluijk & Parker 1986). Post-discovery reserves estimates do not fare any better. According to a recent study by Dromgoole & Speers (1997), field reserves on the UKCS, from the time of Government sanction (of development plan) to 4 years after production, were over- or underestimated by 10±20% depending on reservoir complexity. There is little doubt that had the period covered been longer, the error would have been larger. In the Norwegian sector of the North Sea, also, field reserves after discovery show considerable fluctuation through time, with initial estimates being generally on the conservative side (Hermanrud et al. 1996). Ble¬haut (1991) reported similar conservatism from East Malaysia. Early conservatism in reserves estimates was also the hallmark of many giant oil fields, e.g. East Texas, USA (Arps 1956). Unpublished studies tell a similar story. A review of field data by a major oil company several years ago revealed significant underestimation of reserves after discovery: over a period of 9 years, the reserves had increased by 60%. The aforementioned statistics refer to the `best' or `representative' estimate of reserves, e.g. the expectation value, which is a single-value estimate. Reserves estimation, however, should encompass not only the best estimate, but also the uncertainty surrounding the best estimate. Unfortunately, few companies seem to pay attention to how reserves uncertainty changes through time. As it will be noted later, reserves uncertainty has a significant impact on upstream economics. The focus of this paper is to investigate historical trends in reserves uncertainty for a number of fields, interpret the results, and discuss ramifications. The analysis is confined to the preproduction phase, when reserves estimates are based on well data and seismic (and perhaps analogue cases), to the exclusion of well performance data. The historical trends therefore reflect primarily the impact of seismic and appraisal on reserves uncertainty. The pre-production phase is also the phase when all the major investment decisions (facilities and infrastructure) for field development are made.