INTRODUCTION TO WELL LOGS And BAYES THEOREM

Transcription

1 INTRODUCTION TO WELL LOGS And BAYES THEOREM EECS 833, 7 February 006 Geoff Bohling Assistant Scientist Kansas Geological Survey geoff@kgs.ku.edu Overheads and resources available at 1

2 Critical etroleum Reservoir roperties etroleum geologists and engineers are interested in the properties of the rocks constituting oil and gas reservoirs. In particular, they are interested in characterizing the distribution of the following properties in the subsurface: orosity: The volumetric proportion of void space in a given volume of rock; includes a range of pore sizes from mm ( microns) in sedimentary rocks. ore size is generally a function of grain size (larger grains, larger pores) and occurs in a variety of ways, including spaces between grains and/or crystals, as molds of dissolved grains, or as vugs, which are large pores resulting from dissolution of a portion of the rock. The pore space is filled with fluids, including water, oil, and gas. orosity is expressed as a percentage or a decimal ratio. ermeability: A property characterizing the ease with which fluids flow through the rock; dependent on both porosity and pore space geometry. ermeability is generally expressed in millidarcies (md). Relative Saturations: The proportion of pore space occupied by water, oil, and gas are also important, since these influence the overall volume of oil or gas in the reservoir and the relative permeability of the formation to each fluid. (Oil cannot flow through a water-filled pore.) Facies: Facies is a rather vague term that in general means rock type. It is vague because rocks can be categorized

3 in different ways depending on the goals of the analysis. etroleum geologists are generally interested in categorizing rocks according to their ability to store and transmit fluids that is, their porosity and permeability. For example, sandstones generally have relatively high porosity and permeability, while shales have high porosity but low permeability. The application that Marty Dubois will describe involves carbonate rocks (limestones and dolomites), which can display a wide range of porosities and permeabilities depending on how they were formed and their history subsequent to original deposition. The facies distribution is a first order control on the distribution of porosity and permeability in a reservoir. Geophysical (Wireline) Well Logs Many wells drilled for petroleum exploration are logged, meaning that a tool (or tools) containing various sensors is lowered into the borehole and then drawn back up while the sensors measure various properties of the surrounding rock. The measured properties generally include electrical, acoustic, and nuclear properties of the surrounding medium the combined properties of both the rock matrix and the fluids in the pore space. The resulting records of measured properties versus depth are variously referred to as wireline logs (because the tools are lowered on a wireline), well logs, geophysical logs, or just plain logs. These logs give indirect information regarding the distribution of the critical reservoir properties discussed above and log analysts spend most of their time trying to infer reservoir properties from logs. 3

4 Some of the more common logs include: Gamma Ray: Measures natural radioactivity of the surrounding rock. The most abundant radioactive isotopes, those of otassium, Thorium, and Uranium, occur in higher concentrations in the clay minerals that constitute shales, so the gamma ray log is particularly useful for distinguishing shales from other rocks. Density and Neutron orosity: Two different sensors providing estimates of the porosity, the first based on the bulk density of the surrounding rock and the second based on neutron-absorbing capability of the medium. These two porosity estimates tend to have complementary biases, so they are often combined to get a more accurate estimate. Electrical Resistivity: Because of the large contrast in electrical resistivity between most rock-forming minerals (highly resistive) and any contained fluids (more conductive), variations in bulk electrical resistivity are dictated primarily by variations in porosity and in the relative saturations of the fluids in the pore space. Given a porosity estimate from one or both of the density logs described above, the resistivity log is generally used to estimate the relative saturations of water, oil, and gas. hotoelectric factor: Measures the photoelectric absorption capacity of the surrounding medium; sensitive to the mineralogical composition of the rocks and thus good for discriminating lithology (rock type sensu stricto). 4

5 In somewhat idealized form, a set of logs versus depth in the borehole (in feet the numbers in the center track ) might look like the following. This is actually a synthetic example, with logs computed from a perfectly known sequence of rock types (drawn in the depth track) and with no influence from fluid variations, but it is fairly realistic: For more information on log analysis, see General reference on petroleum geology: Selley, Richard C., Elements of etroleum Geology, Second Edition, 1998, Academic ress, San Diego. 5

6 Development of Bayes Theorem Terminology: (A): robability of occurrence of event A (marginal) (B): robability of occurrence of event B (marginal) (A,B): robability of simultaneous occurrence of events A and B (joint) (A B): robability of occurrence of A given that B has occurred (conditional) (B A): robability of occurrence of B given that A has occurred (conditional) Relationship of joint probability to conditional and marginal probabilities: ( A, B) ( A B) ( B) or ( A, B) = ( B A) ( A) = So... (A B)(B) = (B A)(A) 6

7 Rearranging gives simplest statement of Bayes theorem: ( B A) = ( A B) ( B) ( A) Often, B represents an underlying model or hypothesis and A represents observable consequences or data, so Bayes theorem can be written schematically as ( model data) ( data model) ( model) This lets us turn a statement about the forward problem: (data model): probability of obtaining observed data given certain model into statements about the corresponding inverse problem: (model data): probability that certain model gave rise to observed data as long as we are willing to make some guesses about the probability of occurrence of that model, (model), prior to taking the data into account. 7

8 Or graphically, Bayes theorem lets us turn information about the probability of different effects from each possible cause: into information about the probable cause given the observed effects: (Illustration styled after Sivia, 1996, Figure 1.1) 8

9 Assume that B i represents one of n possible mutually exclusive events and that the conditional probability for the occurrence of A given that B i has occurred is (A B i ). In this case, the total probability for the occurrence of A is ( A) n = i= 1 ( A B ) ( ) i B i and the conditional probability that event B i has occurred given that event A has been observed to occur is given by ( B A) ( A B ) ( B ) i i i = = n j=1 ( A B ) ( B ) j j ( AB ) ( B ) i ( A) i. That is, if we assume that event A arises with probability (A B i ), from each of the underlying states B i, i=1,,n, we can use our observation of the occurrence of A to update our a priori assessment of the probability of occurrence of each state, (B i ), to an improved a posteriori estimate, (B i A). 9

10 Discrete-robability Example: Dolomite/Shale Discrimination Using Gamma Ray Log Threshold Reservoir with dolomite pay zones and shale non-pay zones. Gamma ray log: Measures natural radioactivity of rock; measured in AI units Shales: Typically high gamma ray (~110 AI units) due to abundance of radioactive isotopes in clay minerals; somewhat lower in this reservoir (~80 AI units) due to high silt content Dolomite: Typically low gamma ray (~10-15 AI units), but some hot intervals due to uranium Can characterize gamma ray distribution for each lithology based on core samples from wells in field: Dolomite Shale Mean Std. Dev Count Cores are sections of rock extracted from the borehole during the drilling process. Cores provide the most direct information on many reservoir properties, but they are expensive. Their utility is often extended by calibrating log responses to core properties in cored wells and then using the derived relationships to predict those properties from logs in uncored wells. 10

11 Gamma ray distributions for dolomite and shale 0.04 dolomite robability Density shale Gamma Ray (AI Units) Will use these distributions to predict lithology from gamma ray in uncored wells, first using a simple rule: - if GammaRay > 60, call the logged interval a shale - if GammaRay < 60, call it a dolomite 11

12 Using Bayes rule we can determine the posterior probability of occurrence of dolomite and shale given that we have actually observed a gamma ray value greater than 60. Let s define events & probabilities as follows: A: GammaRay > 60 B 1 : occurrence of dolomite B : occurrence of shale (B 1 ): prior probability for dolomite based on overall prevalence 60% (476 of 771 core samples) (B ): prior probability for shale based on overall prevalence 40% (95 of 771 core samples) (A B 1 ): probability of GammaRay > 60 in a dolomite = 7% (34 of 476 dolomite samples) (A B ): probability of GammaRay > 60 in a shale = 95% (80 of 95 shale samples) Then the denominator in Bayes theorem, the total probability of A, is given by ( A) = ( A B ) ( B ) + ( A B ) ( B ) 1 1 = 0.07* * 0.40 = 0.4 1

13 If we measure a gamma ray value greater than 60 at a certain depth in a well, then the probability that we are logging a dolomite interval is ( B A) ( A B ) ( B ) ( A) 0.07* = = = 0.10 and the probability that we are logging a shale interval is ( B A) ( AB ) ( B ) ( A) 0.95* = = = Thus, our observation of a high gamma ray value has changed our assessment of the probabilities of occurrence of dolomite and shale from 60% and 40%, based on our prior estimates of overall prevalence, to 10% and 90%. We can do simple sensitivity analysis with respect to prior probabilities. For example, if we take prior probability for shale to be 0% (meaning prior for dolomite is 80%), then get posterior probability of 77% for shale (3% for dolomite) if the gamma ray value is greater than 60 AI units. 13

14 Continuous-robability Example: Dolomite/Shale Discrimination Using Gamma Ray Density Functions It is also possible to formulate Bayes theorem using probability density functions in place of the discrete probabilities (A B i ). We could represent the probability density function that a continuous variable, X, follows in each case as f(x B i ) or, more compactly, f i (x). Then ( B x) i = n f j= 1 i f ( x) ( B ) j ( x) ( B ) i j. That is, if we can characterize the distribution of X for each category, B i, we can use the above equation to compute the probability that event B i has occurred given that the observed value of X is x. For example, based on the observed distribution of gamma ray values for dolomites and shales, a gamma ray measurement of 110 AI units almost certainly arises from a shale interval, because the probability density function for gamma ray in dolomites evaluated at 110 AI units, f 1 (x=110), is essentially 0. This form of Bayes theorem lets us develop a continuous mapping from gamma ray value to posterior probability. 14

15 Shale/Dolomite Discrimination Using Normal Density Functions Dolomite (1) Shale () Mean ( x ) Std. Dev. (s) Count f f 1 [ s ] ( x) = exp ( x x ) 1 1 s1 π 1 [ s ] ( x) = exp ( x x ) s π 1 Normal Approximations for Gamma Ray Distributions 0.04 Kernel density estimate Normal density estimate 0.03 robability Density 0.0 dolomite shale Gamma Ray (AI Units) 15

16 Let Let q = (B ) represent prior probability for shale prior for dolomite is then (B 1 ) = 1 - q p (x) = (B x) represent posterior probability for shale posterior for dolomite is then (B 1 x) = 1 - p (x) So, posterior probability for shale given that the observed gamma ray value = x is p ( x) = q f( x) ( 1 q ) f ( x) + q f ( x) 1 Shale Occurrence robability Using Normal Densities osterior robability for Shale (solid lines) normal pdf for dolomite prior probability for shale used to compute posterior normal pdf for shale robability Density (dashed lines) Gamma Ray (AI Units)

17 Bayes rule allocation: Assign observation to class with highest posterior probability. For base case prior of 40% for shale, 50% posterior probability point occurs at gamma ray of 59.6 so Bayes rule allocation leads to basically same results as thresholding at 60 AI units. But now have means for converting gamma ray to continuous shale probability log. 17

18 Shale/Dolomite Discrimination Using Kernel Density Estimates No need to restrict approach to just normal densities. Could use any other form of probability density function for each category, including the kernel density estimates shown initially: Shale Occurrence robability Using Kernel Densities 0.05 osterior robability for Shale (solid lines) kernel pdf for sandstone prior probability for shale used 0. to compute posterior kernel pdf for shale robability Density (dashed lines) Gamma Ray (AI Units) 0.00 Kernel density estimates are essentially smoothed histograms, scaled to represent legitimate probability density functions (non-negative everywhere and integrating to 1). They can be computed using the ksdensity function in Matlab s statistics package. 18

19 Relationship to Discriminant Analysis Could just as easily use multivariate density functions in Bayes theorem. For example, could be discriminating facies based on a vector of log measurements, x, rather than a single log. If use multivariate normal density functions for each class, Bayes rule allocation leads to classical discriminant analysis. Assuming covariance matrices all equal for different classes leads to linear discriminant analysis: Bayes rule allocation draws linear boundaries between classes in x space. Assuming unequal covariance matrices leads to quadratic discriminant analysis: Bayes rule allocation draws quadratic boundaries between classes. We will talk more about discriminant analysis next time. 19

20 An Excel workbook implementing the shale/dolomite discrimination example is available at 0

21 1