Publication History: This article is based on "Crain's Analyzing Unconventional Reservoirs" by E. R. (Ross) Crain, P.Eng., first published in 2006, and updated annually until 2016. This webpage version is the copyrighted intellectual property of the author.

Do not copy or distribute in any form without explicit permission.

Igneous and Metamorphic Basics


 Oil and gas in igneous and metamorphic rocks is not that uncommon. There are good examples in Viet Nam, Indonesia, Japan, Venezuela, and even the USA, where they are called “bald highs”.

To be economically attractive, these unconventional reservoirs need all the same attributes as a sedimentary play. They must have a trap with a seal and porosity to hold fluids. There must be a source of hydrocarbons, a migration path, and migration must have taken place. The porosity is usually related to natural fractures augmented by hydrothermal solution porosity formed near the fracture surfaces prior to hydrocarbon migration. Solution porosity will not form after the oil or gas is in place.

The source and migration path are often from obvious normal sediments near or surrounding the igneous or metamorphic rocks. Russian, American, and Canadian scientists have postulated different deep-seated, abiotic, non-sedimentary sources for oil in granite. Most igneous/metamorphic rocks with oil are updip from somewhere, so you don’t NEED deep genesis to get oil into them. But you do need pore space to hold it. This infers hydrothermal action to generate that porosity, so if water can get there, why not conventional oil or gas? If oil, methane, helium, and hydrogen can percolate up through miles of mantle and crust, why does it stop in the granite and not keep going to the surface? ….. George Noori, host of the all-nigh talk-radio show “Coast to Coast AM”, believes in the infinite abiotic oil theory. Who can argue with George?

From a petrophysical point of view, it doesn’t matter how the oil or gas got there. Oil is where you find it and we analyze these wells the same way as any other, taking care to recognize the unique mineralogy and the low porosity. For example, White Tiger (Bach Ho) Field in Viet Nam averages only 1.7 percent porosity. In addition, most igneous and metamorphic reservoirs are fractured to some degree. This section is a brief review of fractured reservoirs.

Fracture Porosity Definitions showing the distinction between fracture porosity and fracture-related porosity (solution porosity).

Fracture porosity is usually very small. Values between 0.0001 and 0.001 of rock volume are typical (0.01% to 0.1%). Exceptions certainly exist, but only where rapid cooling caused significant contraction . Fracture-related porosity caused by hydrothermal solution or surface weathering may attain much larger values, but the porosity in the actual fracture is still very small.

Fracture porosity is found accurately only by processing the formation micro-scanner curves for fracture aperture and fracture frequency (fracture intensity). All other methods, including the well known “dual-porosity” model, are extremely inaccurate. These models either over-estimate fracture porosity by several orders of magnitude, or cannot be applied because the log data does not fit the model.

The effect of fracture porosity on reservoir performance, however, is very large due to its enormous contribution to permeability. As a result, naturally fractured reservoirs behave differently than un-fractured reservoirs with similar porosity, due to the relative high flow capacity of the secondary porosity system. This provides high initial production rates, which can lead to extremely optimistic production forecasts and sometimes, economic failures when the small reservoir volume is not properly taken into account.

The only correct approach is to use formation micro-scanner fracture aperture and frequency data:

1: PHIfrac = 0.001 * Wf * Df * KF1
            2: Kfrac = 833 * 10^11 * PHIfrac^3 / (Df^2 * KF1^2)

3: Kfrac = 833 * 10^5 * PHIfrac * Wf^2

4: Kfrac = 833 * 10^2 * Wf^3 * Df * KF1


  KF1 = number of main fracture directions

         = 1 for sub-horizontal or sub-vertical

         = 2 for orthogonal sub-vertical

                     = 3 for chaotic or brecciated

  PHIfrac = fracture porosity (fractional)

  Df = fracture frequency (fractures per meter)

  Wf = fracture aperture (millimeters)
  Kfrac = fracture permeability (md)


Kfrac can be many thousands of millidarcies. Equations 2, 3, 4 give identical results.

To be an economic success, fractures must be connected to a reasonable hydrocarbon bearing reservoir with sufficient volume to warrant exploitation. Fractures alone are seldom enough. If there is no reservoir volume, a lot of fractures won’t help much unless there is sufficient fracture related solution porosity to hold an economic reserve.


This can be determined by normal log analysis techniques. Although both the presence of fractures and the presence of a reservoir can be determined from logs, a production test will be needed to determine whether economic production is possible. The test must be analyzed carefully to avoid over optimistic predictions based on the flush production rates associated with the fracture system. Local correlations between fracture intensity observed on logs and production rate are also used to predict well quality.

Resistivity image log in fractured granite reservoir showing computed dip angle and direction

Metamorphic rock classification
Metamorphic rocks
are conventional sedimentary rocks that have been exposed to high heat and pressure. There are several types of metamorphism:

  • Contact metamorphism - changes in the rock due to heat from nearby magma.

  • Regional metamorphism - causes change through intense heat and pressure.

  • Hydrothermal metamorphism - chemical changes in the rock due to the circulation of hot liquids through the rock fractures.

  • Fault zone metamorphism - metamorphic changes caused friction at fault movements.

The quality of the rock is based on the amount of heat and pressure applied to it during the metamorphic processes. Changes that occur during metamorphism are:

  • Re-crystallization - occurs when small crystals join together to create larger crystals of the same mineral.
  • Neomorphism - new minerals are created from the original mineral composition.
  • Metamorphism - new minerals are created by gaining or losing chemicals.

Specific sedimentary rocks become specific metamorphic rocks, as shown below:

Sandstone Quartzite
Limestone, Dolomite Marble
Basalt Schist or Amphibolite
Shale Slate
Granite Schist
Rhyolite Schist

The petrophysical properties of metamorphic rocks are often similar to their pre-metamorphic sedimentary counterparts as long as new or different minerals have not formed. Standard 2- and 3-mineral models, or probabilistic models, are used to calculate lithology. The density neutron complex lithology model is used to calculate porosity when data and borehole conditions permit. Sonic neutron crossplot models can be used as an alternate when needed. Shale (or schist) corrections based on a gamma ray log may be needed. Water saturation from Archie is usually sufficient  since schist is not overly conductive. A table of rock properties for the mineral models is shown below.







2.65  2.60  2.54--2.72 2.70  2.68

55.0  55.1  46.2 53.5  54.5 101.2  92.4 101.2 0.0 - 0.015

1.82 / 4.82

Gneiss 2.73  2.64--2.91 54.2  38.7 105.1 121.1    
Gneiss dark 3.04  2.83 54.7  45.5 100.8  96.7    
Gneiss pink/grey 2.75  2.80 48.6  48.3 99.0  94.3    
Greywacke 2.705 52.9  49.2--56.4 81.2    
Marble 2.736  2.58--2.93  2.71 52.5  41.6  57.1
95.9  107.7    
Marble Dolomite 2.825  2.825 43.5  43.5   0.0  
Marble Calcite 2.77  2.77 49.8  49.5   0.0  
Monzonite 2.64  2.63--2.65 56.5  52.4--59.8 96.1    
Phyllite 2.79  2.89 58.5  47.5   0.08  
Schist Mica 2.79  2.83 60.0  49.5   0.065  

These tables are in English units. If you work in Metric units, multiply density by 1000, and multiply sonic by 3.281. Data includes averages and ranges from lab measurements and values from log crossplots. Some porosity and variable mineralogy effects are probably present. Source: "Matrix Values for Sonic, Density, and Neutron", SSI, 1980 in SPWLA Geothermal Handbook 1982.

The rock property variations suggest that the mineral makeup and porosity may vary a bit between samples. Some x-ray diffraction (XRD) data may help sort out the mineral composition more precisely in the real world of oil field petrophysical analysis.

Igneous rock classification
Igneous rocks are classified in several ways – by composition, texture, and method of emplacement. The composition (mineral mixture and internal porosity) determines the log response. The texture determines the name used for the mineral mixture, and the method of emplacement determines the texture and internal porosity structure (if any). The same mineral mixture can have more than one name based on its crystal size and method of emplacement.

Intrusive igneous rocks are formed inside the earth. This type of igneous rock cools very slowly and is produced by magma from the interior of the earth. They have large grains, may contain gas pockets, and usually have a high fraction of silicate minerals. Intrusions are called sills when lying roughly horizontal and dikes when near vertical.

Extrusive igneous rocks form on the surface of the earth from lava flows. These cool quickly. They have small grains and contain little to no gas.

Both intrusive and extrusive rocks may contain natural fractures from contraction while cooling, and may have carried non-igneous rocks with them, called xenoliths.

Intrusive rocks may alter the rocks above and below them by metamorphosing (baking) the rock near the intrusion. Extrusives only heat the rock below them, and may not cause much alteration due to rapid cooling. Extrusives can be buried by later sedimentation, and are difficult to distinguish from intrusives, except by their chemical composition and grain size.

The mineral composition of an igneous rock depends on where and how the rock was formed. Magmas around the world have different mineral make up.

Granite Wash (yellow) and shale (gray) above granite (white) and diorite (tan). Porosity and lithology are calculated from conventional density, neutron, and PE methods. Both zones are radioactive so the GR is not used for shale volume.

Felsic igneous rocks are light in color and are mostly made up of feldspars and silicates. Common minerals found in felsic rock include quartz, plagioclase feldspar, potassium feldspar (orthoclase), and muscovite. They may contain up to 15% mafic mineral crystals and have a low density.

Mafic igneous rocks are dark colored and consist mainly of magnesium and iron. Common minerals found in mafic rocks include olivine, pyroxene, amphibole, and biotite. They contain about 46-85% mafic mineral crystals and have a high density.

Ultramafic igneous rocks are very dark colored and contain higher amounts of the same common minerals as mafic rocks, about 86-100% mafic mineral crystals.

Intermediate igneous rocks are between light and dark colored. They share minerals with both felsic and mafic rocks. They contain 15 to 45% mafic minerals.

Plutonic and volcanic rocks generally have very low porosity and permeability. Natural fractures may enhance porosity by allowing solution of feldspar grains. Some examples with average porosity as high as 17% are known.

Diorite intrusion covered by shale, open fracture visible as low resistivity spike at top of diorite. Fracture aperture (Track 2) and fracture porosity (Track 5) are from resistivity image processing.

Tuffs and tuffaceous rocks have high total porosity because of vugs or vesicles in a glassy matrix. This is most common in pyroclastic deposits. Interparticle porosity may also exist. Some effort has to be made to separate ineffective microporosity from the total porosity. Pumice (a form of tuff) has enough ineffective porosity to allow the rock to float! When other minerals fill the vesicles by precipitation, the tuff is called a zeolite. 






Gamma Ray

Coarse Crystalline

Fine Crystalline


Silica Content








Rhyolite Tuff




Dacite Tuff




Andesite Tuff




Zeolite Tuff











For quick-look identification of igneous rocks, crossplots have been widely used for many years. Before the advent of the PE curve, crossplots using neutron, sonic and density were the best bet. Some prior calculations are required. Matrix density (DENSma), sonic matrix travel time (DTCma), lithology factors Mlith and Nlith must be derived. With the PE curve, a lithology factor called Plith can be added, as well as Uma, the matrix capture cross section. These calculations are covered in other Chapters of this Handbook. Examples are shown below.

DENSma vs DTCma Crossplot

Mlith vs Nlith Crossplot

Igneous rock properties
Petrophysical properties of igneous rocks are quite variable because mineral composition can be quite variable and poorly defined. Standard 2- and 3-mineral models are used to calculate lithology. The density neutron complex lithology model is used to calculate porosity when data and borehole conditions permit. Sonic neutron crossplot models can be used as an alternate when needed. Water saturation from Archie is the normal method.
A table of rock properties for the mineral models is shown below.







2.65  2.60  2.54--2.72 2.70  2.68

55.0  55.1  46.2 53.5  54.5 101.2  92.4 101.2 0.0 - 0.015

1.82 / 4.82


2.65  2.667  2.52--2.81 2.65  2.68  2.45--2.94
2.88  2.62

50.0  50.8  46.8--53.5  64.9  53.5 56.2  42.6  42.6  55.1 85.7  104.3 82.7  82.4 0.01

2.70 / 7.00



50.0 87.8--96.5  93.2 0.0


    Rhyolite Tuff

2.50  2.39  2.05  2.258

55.0  74.3  72.9
97.1  123.9 153.9 131.4 0.04



2.72  2.63

55.0  69.3 97.1 0.02

3.25 / 8.75



60.0 89.9 0.02



2.81 2.88  2.80--2.93

57.0  57.5  55.4--59.8 89.9 0.035

3.56 / 9.91

    Andesite and Andesite Breccia

2.76  2.68  2.70  2.67 2.62--2.70  2.73  2.73 2.32  2.57

58.0  56.3  53.8
56.1--56.4  53.4--54.2  56.3  53.8--58.3  64.0 60.7--66.0  64.0  55.8 54.5--57.3  57.3
91.5  94.7 98.5  96.1 96.1  114.6 112.5--121.9 176.2 0.18



2.85  2.97  2.91--3.02 2.91

57.1  57.1  57.1  55.0-59.8 96.8  96.8 94.1 0.04



3.05  2.55  3.03  2.97  2.69  2.73

47.8  61.1  47.0--54.0 47.0  56.8  56.3--57.3 54.8  54.6--55.0 101.9  110.9 85.1  99.3 103.3 0.147


    Zeolite Tuff

2.30  2.58

60.0  63.5 159.7 0.20



2.94  2.976  2.85--3.12 2.94  2.95  2.80--3.05

42.4 42 .4 42.2--47.6 43.7  47.6  44.3--52.6 90.1  90.1 0.05

4.80 / 13.3



44.6 85.8  


    Diabase Lava


55.0   0.24



3.29  3.30  3.24--3.74 3.29  3.30  2.96--3.76 3.25

38.2  38.2 34.7--41.1 37.9  42.2  36.7--55.8 48.0  45.8--51.8 76.9  76.9 82.6 0.04


Anorthosite 2.74  2.68--2.81  2.75 47.5  45.3--53.5  45.4 45.3  44.8--46.2 84.3  82.6    
Basalt Olivine 3.00  2.83 51.5  51.4--51.6  53.1 52.3--53.8 85.9  89.9    
Granite Conglomerate 2.75  2.78 57.2  58.4      
Diabase 2.98  2.96--3.05  2.95 2.998  2.97 44.6  44.0 -46.0  43.8 47.7  38.9--59.2  51.1 50.8--51.7 85.8  90.2    
Norite 2.984  2.72--3.02  2.97 2.96  2.85--3.08 44.1  43.5--49.0  47.2 46.2--49.3 89.8    
Obsidian 2.35  2.35  2.39
52.4  52.1--52.5  52.1 52.5  52.4--52.5 85.4  85.9    
Phyllite 2.79  2.89 58.5  47.5   0.08  
Tuff Undescribed 1.38 213.1  201.8--224.1 350.3    

Tuff Zeolite

2.30  2.58

60.0  63.5   0.20


Tuff Ash Flow 2.69 62.8      
Tuff Welded 2.56 59.8 146.8    

Tuff Rhyolite

2.50  2.39  2.05  2.258

55.0  74.3  72.9
97.1  123.9 153.9 131.4 0.04


Volcanic Brecchia 2.19 72.2  72.2 122.4    

Igneous rock properties vary a lot between samples due no doubt to varying mineral composition and inadequate description of the rock type. All these values have a moderate range (+/- 10%) and some tuning may be necessary.

These tables are in English units. If you work in Metric units, multiply density by 1000, and multiply sonic by 3.281. Data includes averages and ranges from lab measurements and values from log crossplots. Some porosity and variable mineralogy effects are probably present. Source: "Matrix Values for Sonic, Density, and Neutron", SSI, 1980 in SPWLA Geothermal Handbook 1982.

Igneous MINERAL properties
Properties for individual minerals are better known and less variable than rock-type values, since logs respond only to minerals and fluids and not grain size. It is more accurate to use a mineral model than a rock-type model. Here are the mineral properties that can be used in the various multi-mineral log analysis models. These are the same values that might be used in a sedimentary rock sequence, sorted to reflect the common constituents of igneous rocks. Minerals that appear in small quantities are not listed here.


































K Feldspar
































The mineral properties in the above table have been used successfully in igneous rock environments but the numbers may need to be tuned for each situation. Use these values in the standard two and three mineral , or probabilistic, models to determine mineral composition. The results are then compared to igneous rock-type descriptions to solve for rock-type. See next Section for details on how this is done. Values for additional minerals can be found HERE.

Since a typical log suite can solve for 3 or 4 minerals at best, you need to choose the dominant minerals and zone your work carefully. If you have additional useful log curves, you might try for more minerals or set up several 4 mineral models in a probabilistic solution. A good core or sample description will help you choose a reasonable mineral suite.

Sometimes a mineral is determined by triggers. For example, where basalt beds are interspersed between conventional granites or quartzites, it is easy to use the PE or density logs to trigger basalt, leaving the remaining minerals to be defined by a two or three mineral model. This approach is widely used in sedimentary sequences to trigger anhydrite, coal, or salt.

The illustration below gives the average mineral composition of some igneous rock-types.

Typical mineral composition of igneous rocks. 

After determining the mineral composition, the rock-type can be estimated from a ternary diagram or by a near-fit to the mineral composition shown in the above illustration. The best mineral model is usually plagioclase + quartz + K feldspar. If quartz is very low, orthopyroxene should be used instead.




















K Feldspar






Orthopyroxene 0.00 0.00 0.00 0.00 0.15
Other 0.08 0.18 0.20 0.32 0.16


Since a typical log suite can solve for 3 or 4 minerals at best, you need to chose the dominant minerals and zone your work carefully. If you have additional useful log curves, you might try for more minerals or set up several 4 mineral models in a probabilistic solution. A good core or sample description will help you choose a reasonable mineral suite.


Here is a granite/metamorphic example from Indonesia. The reservoir has a porous granite at the base, metamorphic sandstone above, topped by conventional sandstone. Porosity is moderately low throughout but the gas column is continuous. Interbedded shales (schist or gneiss in the metamorphic interval) are present but do not act as barriers to vertical flow.

In this case, the mineralogy was calibrated by quantitative sample descriptions, which in turn were keyed to raw log response to minimize cavings and depth control issues. Porosity and water saturation were derived from conventional log analysis methods. The reservoir is naturally fractured and a fracture intensity curve was generated from anomalies on the open hole logs. This was compared to the fracture intensity from resistivity micro image log data.

Metamorphic / Granite example with spectral GR (Track 1), total gas (Track 2), resistivity (Track 3), fracture aperture, fracture intensity, fracture porosity (from FMI processing, Track 4), density, neutron, PE (Track 5), log analysis porosity (Track 6), water saturation (Track 7), core permeability (Track 8), quantitative sample description (Track 9), calculated lithology (Track 10).

Compare fracture intensity from log anomalies (black shaded curve in porosity track with fracture intensity from FMI (red curve, track 4). Best gas production in granite is confirmed by gas show on gas log and by production logging in open hole. Sample descriptions show minerals as seen in microscope (quartz, feldspar, mica) to nearest 5%. Log analysis lithology show rock type, not minerals (quartz, granite, granodiorite). The sands and shale immediately above the granite are metamorphosed, visible in samples, but there is little effect on log properties except for low clay bound water on neutron and density logs in the shale/slate. Some wells had limestone marble in the metamorphosed interval.

Most people forget that there are many unconventional reservoirs in the world, including igneous, metamorphic, and volcanic rocks. Granite reservoirs are prolific in Viet Nam, Libya, and Indonesia. Lesser known granite reservoirs exist in Venezuela, United States, Russia, and elsewhere. Indonesia is blessed with a combination sedimentary, metamorphic, and granite reservoir with a single gas leg. Japan boasts a variety of volcanic reservoirs.

This example is from the Bach Ho (White Tiger) Field in Viet Nam.

Log analysis in these reservoirs requires good geological input as to mineralogy, oil or gas shows, and porosity. A good coring and sample description program is essential, and production tests are essential. The analyst often has to separate ineffective (disconnected vugs) from effective porosity and account for fracture porosity and permeability. All the usual mineral identification crossplots are useful but the mineral mix may be very different than normal reservoirs. Many such reservoirs seem to have no water zone and most have unusual electrical properties (A, M, N), so capillary pressure data is usually needed to calibrate water saturation.

Ternary Diagram for Granite 

In the example below, the granitic mineral assemblage was defined by the ternary diagram at right. The three minerals (quartz, feldspar, and plagioclase) were computed from a modified Mlith vs Nlith model, in which PE was substituted for PHIN in the Nlith equation. If data fell too far outside the triangle, mica was exchanged for the quartz.

Three rock types, granite, diorite, and monzonite, were derived from the three minerals. A trigger was set to detect basalt intrusions. A sample crossplot below shows how the lithology model effectively separates the minerals.

Mlith vs Plith crossplot for granite (micaceous data excluded)

In this fractured granite example, raw data curves are shown in Tracks 1, 2, and 3 with effective porosity, water saturation, and matrix permeability in Tracks 4, 5, and 6. The mineral model calculated from the log analysis is in Track 7 and the rock type model calculated from the minerals using the ternary diagram is in Track 8. Basalt was triggered from high density or high PE or both.

A sample of the log analysis plot is shown above. The average porosity from core and logs is only 0.018 (1.8%) and matrix permeability is only 0.05 md. However, solution porosity related to fractures can reach 17% and permeability can easily reach higher than several Darcies. Customized formulae were devised to estimate these properties from logs, based on core and test data. My colleague Bill Clow devised most of the methods used on this project.

Fracture porosity from resistivity micro scanner logs was also computed where available to help control the open hole work. A black and white resistivity image log below shows some of the fractures. Both high and low angle fractures co-exist.

Resistivity micro scanner image in granite reservoir

It is clear that non-conventional reservoirs may need some extra effort, customized models, and unique presentations. Everything you need to develop these techniques can be found elsewhere in this Handbook. The mineral properties need to be chosen carefully, but the mathematical models don't change too much.

Except for granite wash reservoirs, and maybe some weathered granite associated with them, there are is no known granite production in Canada. But that doesn't stop people from looking and finding hints of gas. This well was purposely positioned to evaluate granite potential, with emphasis on the possibility that gas is generated  deep in the mantle, and percolates up to shallower porous intervals. With fractures and some porosity from solution of the granite, gas could be trapped. Below is a portion of the petrophysical analysis, run with a multi-mineral statistical model, instead of a deterministic model as in the previous examples.

Top page of four showing a granite test in Canada near Fort McMurray. Log data was processed with a statistical multi-mineral model that included quartz, feldspar, plagioclase, and water. This portion of the granite is slightly porous, lightly fractured, and mostly wet. Deeper portions showed lower porosity but calculate some gas content, also seen on the gas log. However open hole tests produced salt water. Curves shown are spectral GR (Track1), resistivity (20 to 200000, Track 2), density, neutron, PE (Track 3), water saturation (Track 4), porosity (Track 5), Permeability (Track 6), lithology (quartz, feldspar, plagioclase - Track 7).

Page Views ---- Since 01 Jan 2015
Copyright 1978 - 2018 E. R. Crain, P.Eng. All Rights Reserved