P. 1
Zolotukhin - Reservoir Engineering

Zolotukhin - Reservoir Engineering


|Views: 3,295|Likes:
Published by agmyatkyaw77
Great book, everythings about RE are introduced here.
Great book, everythings about RE are introduced here.

More info:

Published by: agmyatkyaw77 on Dec 22, 2008
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





The methods of reserve estimation based on reservoir data are volumetric and can be divided
into deterministic and probabilistic (stocastic) estimates. The main difficulty in a volumetric
estimate of resources/reserves is in the transfer of data obtained at a small scale (core analysis,
lithofacies data, well logs, etc.) into a much larger scale ( i.e. data "upscaling" for interwell

Deterministic Methods

The principle of a deterministic approach to resources/reserve estimates is to "upscale" the
information derived from the wells and supported by seismic survey, into the interwell space
by using an interpolation technique.
The main parameters used for a volumetric estimate in this approach are:

• The reservoir "gross" isopach map, which means the bulk thickness of the reservoir rocks


• The reservoir "net" isopach map, which means the cumulative thickness of the permeable
rock units only. The Net-to-Gross ratio (N/G) is an important parameter indicating the
productive portion of the reservoir.


Chapter 1. Oil and Gas Resources and Reserves

• The reservoir rock porosity (as a volume-based weighed average):

φ = ∑iφiAihi


whereφ is the local porosity,Ai is a subarea andhi is a subthickness (of permeable rock).

• The permeability and net-thickness product (khN) is important for the estimation of well
production capacity:

(khN) = hN


= N




where ki is the local permeability (other symbols as above).

• Volume-based average saturation of water, gas and oil. For example water saturation:

Sw = ∑iSwiφiAihi


Plotting these parameters as contoured maps (isopachs, isoporosity, isopermeability, etc.)
provides the crucial information on their variation and distribution in the reservoir and makes
it possible to evaluate the reservoir pore volume and its fractions saturated with oil and gas
(hydrocarbon volume). The numerical value of hydrocarbon resources/reserve estimate their
represents an outcome of "integrated" map analysis.

Stochastic Methods

An alternative approach is a probabilistic estimation of resources/reserves, which takes more
account of the estimate uncertainty. Stochastic reservoir description is usually based on the pro-
cedure of random-number generator. This numerical technique assumes that the main reservoir
properties (porosity, permeability, N/G, ect.) all have random, possibly normal, frequency dis-
tributions, with the range of values included by core and well-log data. The maximum and
minimum values are specified for each of the reservoir parameters and the random number
generator then "drowns data", so to speak, and then simulates their actual density distribution
in the whole reservoir.

In practice, it is necessary to repeat the stochastic simulation for different "seeds" (initial
boundary values) in order to asses and quantify the actual variation of a given parameter. Each
numerical realisation bears an uncertainty for the reservoir characterisation, where the prob-
abilistic rather than deterministic, is an estimate of resources/reserves. Different realisations
lead to different volumetric estimates, with different probabilities attached. The cumulative
frequency distributions of these estimates, that is used to asses their likelihood will be a very
unclear formulation. See Fig.1.2.
In common usage [8] we have:

– An estimate with 90 % or higher probability is the level regarded as aproven value.

– An estimate with 50 % or higher probability is the level regarded as aproven + probable


1.2 Methods for Resources/Reserve Estimation


Frequency of cumulative







Probability, that a given
value of resources will
be at least as great
as shown





Figure 1.2: Example of stochastic volumetric estimate based on a se-
ries of random-number simulations.

– An estimate with 10 % or higher probability is the level regarded as aproven +probable
+ possible value.

As more information on the reservoir becomes available, the cumulative frequency graph
may change its shape and the uncertainty of our resource/reserve estimates may decrease, see

More generally, the problem of certainty can be considered in terms of "fuzzy" [61], prob-
abilistic and deterministic estimates based on the data available at a particular time, as seen
in Fig.1.4. A comparison of these estimates may be more revealing that each of them is in

At the very early stages of field appraisal, the data are usually too limited for using statis-
tical analysis and, hence, afuzzy estimate of the resources/reserves may be best or only option
[22, 28, 56]. The lack or scarsity of data in such cases is compensated by a subjective assess-
ment of the reservoir characteristics (i.e. the shape of the distribution and the maximum and
minimum values of a given reservoir parameter), Based on the knowledge from other reservoirs
or simply a theoretical guess. A rectangular distribution means no preference and a triangular
distribution means that strong preference distributions are used.
When moredata have been collected and statistical analysis becomes possible, aprobabilis-
estimate can be made. The range in the possible values of the reservoir parameters would
then be narrower, compared to a fuzzy assessment. When the data available are abadundant, a
deterministic estimate can be made based on a well- specified value of a particular parameter
for a particular part (zone, subunit or layer) of the reservoir.

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->