[First Hit]

Datapages, Inc.Print this page

 Click to view article in PDF format.

 

 

A Petrophysical Method to Evaluate Irregularly Gas Saturated Tight Sands Which Have Variable Matrix Properties and Uncertain Water Salinities*

 

Michael Holmes1, Dominic Holmes1, and Antony Holmes1

 

Search and Discovery Article #40673 (2011)

Posted January 11, 2011

 

*Adapted from oral presentation at AAPG International Conference and Exhibition, Calgary, Alberta, Canada, September 12-15, 2010

 

1Digital Formation, Denver, Colorado ([email protected])

 

Abstract

 

A problem in many Rocky Mountain tight gas sandstones is a sequence that is only partially gas saturated, with changing matrix properties combined with variable (and often unpredictable) water salinities. Often it is difficult to distinguish between high resistivity fresh water wet sands, and high resistivity, gas-bearing sands. A standard approach is to make a qualitative judgment based on Previous HitdensityNext Hit/Previous HitneutronNext Hit response – the gas “cross-over” effect. However, if matrix properties are variable, this approach can be misleading, and is at best a qualitative judgment.

 

The methodology presented here is a quantitative assessment of gas saturation by comparing matrix specific Previous HitdensityNext Hit and Previous HitneutronNext Hit responses with Previous HitporosityNext Hit, calculated such that gas effects are minimized. Cross plot Previous HitporosityNext Hit from Previous HitdensityNext Hit/Previous HitneutronNext Hit combination is only minimally affected by gas and by changing matrix properties.

 

Three sets of calculations are made assuming sandstone (bulk density 2.65 gm/cc), calcareous sandstone (bulk density 2.68 gm/cc) and heavily cemented calcareous sandstone, or limestone (bulk density 2.71 gm/cc). Quantified estimates of gas saturation, as “seen” by each Previous HitlogNext Hit, are available for each assumed rock type. Pressure effects on Previous HitporosityNext Hit Previous HitlogNext Hit responses are included in the calculations.

 

Four sets of saturation profiles are now available, one from standard resistivity Previous HitlogNext Hit analysis and three from Previous HitporosityNext Hit Previous HitlogNext Hit analysis assuming different matrix properties. Comparisons among the 4 sets of saturation profiles can be combined with other data, such as mud Previous HitlogNext Hit shows and (if available) core measured matrix densities. Using such comparisons, it is often relatively simple to distinguish between wet intervals and gas-bearing intervals. With increasing assumed grain density, gas saturations calculated from the Previous HitporosityNext Hit logs increase. If gas saturations so defined are unrealistically high, it is an indication that actual grain density is less than assumed grain density.

 

Additionally, if matrix properties are well-defined, it is possible to verify Rw input for resistivity Previous HitlogNext Hit interpretation, and adjust as necessary. It is important to recognize that the Previous HitporosityNext Hit logs, and particularly the density Previous HitlogNext Hit, investigate close to the wellbore, and may well be influenced entirely by the flushed zone. Examples are presented from Rocky Mountain reservoirs, in sequences where the problem of irregular gas saturation in systems with variable Rw is particularly severe.

 

Figures

 

Copyright � AAPG. Serial rights given by author. For all other rights contact author directly.

 

Abstract
Figures
Methodology
Example




















Abstract
Figures
Methodology
Example


















Abstract
Figures
Methodology
Example


















Abstract
Figures
Methodology
Example


















Abstract
Figures
Methodology
Example


















Abstract
Figures
Methodology
Example


















Abstract
Figures
Methodology
Example


















Abstract
Figures
Methodology
Example


















 

 

fig01

Figure 1. Schematic of gas saturation from Previous HitporosityNext Hit Previous HitlogNext Hit analysis (for shallow reservoirs).

fig02

Figure 2. Depth Previous HitlogNext Hit from standard shaley formation resistivity analysis. Track descriptions are as follows: (1) raw wireline Previous HitlogNext Hit; (2) reservoir composition, where yellow indicates matrix, red indicates Previous HitporosityNext Hit, and grey indicates shale; (3) water saturation, core data illustrated by blue and green symbols; (4) bulk volumes where grey indicates shale, brown indicates hydrocarbons, light blue indicates poor quality reservoir (possible mobile water), and dark blue indicates capillary bound water; (5) pay flags where yellow indicates gross reservoir, red indicates net reservoir, and green indicates net pay.

fig03

Figure 3. Interpretation of Previous HitporosityNext Hit vs. resistivity relations to determine parameters for Archie analysis. a (cementation constant) = 1.0, Rw (water resistivity) = 0.1, m (cementation exponent) = 2, and n (saturation exponent) = 1.7.

fig04

Figure 4. Depth Previous HitlogNext Hit showing gas saturation from Previous HitporosityNext Hit Previous HitlogNext Hit analysis. Track descriptions are as follows: (1) reservoir composition, where yellow indicates matrix, red indicates Previous HitporosityNext Hit, and grey indicates shale; (2) water saturation, core data illustrated by blue and green symbols; (3) bulk volumes, where grey indicates shale, brown indicates hydrocarbons, light blue indicates poor quality reservoir (possible mobile water), and dark blue indicates capillary bound water; (4) matrix properties - grain density and matrix travel time; (5) permeability, core data illustrated by black symbols. Tracks 6, 7, and 8 show Previous HitporosityNext Hit and gas saturation from Previous HitporosityNext Hit logs only. Yellow indicates gas saturation from Previous HitporosityNext Hit logs, and blue indicates water saturation from resistivity Previous HitlogNext Hit analysis. Track 6 shows analysis assuming a grain density of 2.65 gm/cc, track 7 shows analysis assuming 2.68 gm/cc, and track 8 shows analysis assuming a grain density of 2.71 gm/cc. Track 9 contains pay flags where yellow indicates gross reservoir, red indicates net reservoir, and green indicates net pay. 9a shows pay flags from resistivity analysis, 9b shows pay flags from Previous HitporosityNext Hit Previous HitlogNext Hit analysis assuming grain density of 2.65 gm/cc, 9c shows pay flags from Previous HitporosityNext Hit Previous HitlogNext Hit analysis assuming grain density of 2.68 gm/cc, and 9d shows pay flags from Previous HitporosityNext Hit Previous HitlogNext Hit analysis assuming grain density of 2.71 gm/cc.

 

Methodology

 

Shaley Formation Resistivity Analysis

 

The interpretation is a standard shaley formation analysis, involving calculations of total Previous HitporosityNext Hit, shale volume, and total water saturation. One of the issues in tight gas sands is the correct choice of matrix lithology. If only a density Previous HitlogNext Hit is used, the correct choice of grain density is crucial. In the event of lack of core data this may be particularly problematic, especially if grain density is variable. By using a Previous HitdensityNext Hit/Previous HitneutronNext Hit cross plot, this restriction is overcome. Fluid density is a required input, and should be estimated from a reasonable assumption of gas saturation close to the wellbore. A Previous HitporosityNext Hit/resistivity cross plot can help in this choice.

 

The second set of analysis involves the calculation of effective Previous HitporosityNext Hit and water saturation (i.e. removing the effects of clay). From these calculations, again based on resistivity interpretations, potential mobile water and/or poor quality reservoir rock can be distinguished from capillary bound water. Lack of light blue color fill on Figure 2, track 4 and Figure 4, track 3 equates to better quality, gas-bearing reservoirs. Example of application of these techniques are described in AAPG Annual Convention, June 2009 “Relationship Between Previous HitPorosityNext Hit and Water Saturation: Methodology to Distinguish Mobile from Capillary Bound Water ” by Michael Holmes, Antony Holmes, Dominic Holmes.

 

Gas Saturation from Previous HitPorosityNext Hit Logs

 

Comparison of Previous HitdensityNext Hit with Previous HitneutronNext Hit logs is used routinely for a qualitative assessment of the presence of gas – the Previous HitdensityNext Hit/Previous HitneutronNext Hit “cross-over” effect. The response is controlled by the concentration of hydrogen in the pore space. Because gas contains less hydrogen than oil or water, apparent neutron Previous HitporosityNext Hit is suppressed. The cross plot (Figure 1) shows the effects of gas. Previous HitPorosityNext Hit values are lithology specific, i.e. raw logging data must be converted to a known (or presumed) lithology. By assuming different lithologies, the cross plot can be solved for gas saturation. The assumption is made that the Previous HitdensityNext Hit and Previous HitneutronNext Hit logs both investigate about the same volume of rock – or, as in the case of tight gas sands, there is a little variation of gas saturation away from the wellbore (minimal or no invasion by drilling fluid). Three different lithologies are assumed in this paper:

 

 

 

Example

 

Standard resistivity interpretation from a Piceance Basin well, NW Colorado, is shown in Figure 2. Interpretation of the Previous HitporosityNext Hit resistivity cross plot (Figure 3) shows input used for saturation determination:

 

Rw = 0.1 Water resistivity

a = 1.0 Cementation constant

m = 2.0 Cementation exponent

n = 1.7 Saturation exponent

 

A summary of the interpretation is as follows:

  • Sands from 4850-5000 ft are either tight (as at 4890-4910 ft) or wet (as at 4920-4940 ft).
  • From 5000-5140 ft the sands are routinely gas bearing and, for tight sands, fair reservoir quality – average core permeability 0.1 mD or above.
  • From 5140-5200 ft, the sands are tight – average core permeability less than 0.01 mD.

 

Figure 4 shows the same interval as Figure 2, with the addition of interpretations of gas saturation from Previous HitporosityNext Hit logs. Examination of the Previous HitporosityNext Hit Previous HitlogTop Sg data shows the following:

  • None of the sands has a grain density of 2.71 gm/cc (calculations assuming this value show unrealistically high gas saturations).
  • The shallower sands (above 5000 ft) appear to be mostly wet, with a grain density of 2.65 gm/cc - or they are gas bearing with a more saline water (lower Rw) than the shallower sands, with a grain density of 2.68. Core measured grain densities would help solve the problem.
  • Sands from 5000-5140 ft are gas bearing with a grain density of 2.68 gm/cc.
  • Sands below 5140 ft are not gas bearing, regardless of assumed grain density.

 

Return to top.