2G. R. COATES, S. MENGER, M. PRAMMER, D. MILLER, NUMARSPE 38736
"litho/porosity crossplots" being derived by semi-empiricaland semi-deterministic methods are strongly lithology de-pendent. Optimal circumstances (i.e., clean, shale free, forma-tions saturated with fresh water, good quality logs, goodknowledge of mineral composition) are prerequisites to deduceaccurate formation porosity values.
The following example shows that under certain, never-theless common circumstances, the conventional approach canyield an incorrect porosity value. The true formation porosityshall be calculated by a combination of neutron porosity andgamma-gamma density log. The density log shows a bulk den-sity of
= 2.32 g/cm
. The log analyst assumes a matrix den-sity
= 2.65 g/cm
and fluid density
= 1.00 g/cm
In a clean, water saturated formation the neutron log wouldread the correct porosity value
= 0.2. Using equation
(refer to Appendix A), an averaged neutron-density porosity
/ 2 (1)
= 0.2. This is the correct porosity for this forma-tion. In a gas bearing zone however,
will decrease propor-tional to the hydrogen index and the relative fraction of thefluids. In the extreme,
could drop close to 0 p.u. due to thevery low proton density in gas, whereas
= 0.309 will readtoo high. In this situation the averaged porosity would be de-termined with
= 0.219, i.e., the porosity is overestimatedabout 10%.
Although here the neutron porosity log should only be usedas a qualitative indicator, log analysts regularly have to inter-pret
quantitatively due to lack of better alternatives.
In shaly formations the situation is more complicated.Since the log response in shale bearing zones is very difficultto predict, porosity interpretation in those formations poses areal challenge. Interpretation in such zones is based, to at leasta certain degree, on assumptions about the clay content of theformation. This can introduce large uncertainties in the results.For example, a clay contents of only 5% at a true formationporosity
= 0.10 results in a
estimation error of about6% for Illite and 9% for Kaolinite.
In addition to those inherent difficulties, conventional po-rosity analysis methods require a multiplicity of good qualitylogs. In sub-optimal hole conditions the neutron-density log,which is the back-bone of most conventional porosity analysismethods, is prone to give erroneous results caused by, for ex-ample, borehole rugosity. Hence, failure of one of the requiredlogs can jeopardize the entire formation evaluation.
For a review of the physical principles of conventional po-rosity logging, the reader is referred to Appendix A.
MRIL Total Porosity
The MRIL total porosity simply counts the number of hydro-gen atoms in the fluid state within the measurement volume.For details see Appendices A and B.
Historically, relaxation components with relaxation timesT
> 3 ms have been considered a measure of effective poros-ity, i.e., the pore space not associated with clay minerals.
Thus, in liquid-filled pores, the number of hydrogen atoms"seen" by MRIL maps directly into porosity, since the hydro-gen density is very close to that under calibration conditions.
The newly introduced C/TP technique allows MRIL toquantify hydrogen with nuclear relaxation times as fast as 0.5ms and as slow as several seconds.
The NMR signal inte-grated over the complete relaxation spectrum is designated as
(named after the sum-symbol
In gas-bearing (non-flushed) formations, MSIG undercallstrue porosity proportional to gas saturation and inversely pro-portional to gas hydrogen index. Additional processing andinterpretation techniques are required to restore the porosityreading.
A New Approach
Recent papers have demonstrated that NMR can be used toprovide lithology-independent total porosity information.
Hence, in contrast to the conventional approach of litho/porosity crossplots, using MRIL data leads to an easierand more reliable approach in estimation of true formationporosity.
Taking into account the MRIL's proven capability to pro-vide highly accurate total porosity values regardless of bore-hole rugosity, pore size, pore contents (i.e., water, oil, gas) orshaliness, a new approach in porosity analysis is proposed.NMR data is to be used to provide reliable porosity informa-tion directly requiring only a single log, which inherently con-tains all information required to satisfactorily determine for-mation total porosity.
The following examples are presented to illustrate thatlithology independent MRIL total porosity can provide accu-rate and more reliable values than conventional porosity logsin mixed mineral formations even under sub-optimal boreholeconditions.
The first example is taken from a well in SouthTexas. The logs show data from a portion of the Wilcox for-mation which, typically, consists of separate lithofacies, fromsand rich to mudstone shales, potentially deposited in six dif-ferent depositional systems.
This well is located in the LaSalle Delta system which Xue and Galloway
identify asshowing “[…] strong dip-dominated sandstone trends.Downdip, several local thicks lie at or near the ancient shelf margin. They reflect shelf-edge delta depositions or shelf margin slump resedimentation of delta facies or both.” Thesemarine environment deposits have also been influenced bywave action and often include shell remnants and resultingcarbonate cementation.