GB2236177A - Determination of core parameters using a gamma-gamma logging technique - Google Patents

Abstract

A method for the determination of density, diameter, lithology parameter and elemental concentrations of core samples by means of the detection of gamma rays arising from both natural sources within said samples and from radiation induced by an external gamma ray source, optionally cesium (Cs), irradiating said samples uses a Cs-beam width greater than the sample, and comprises the use of a gamma ray detector provided with several count-rate windows over two main energy regions, a first low energy region for the detection of Cs induced gamma radiation to determine independently density, diameter and lithology parameter, and a second high energy region for the detection of natural gamma radiation to determine concentrations of said natural sources. <IMAGE>

Description

DETERMINATION OF CORE PARAMETERS USING A GAMMA-GAMMA LOGGING TECHNIQUE The present invention relates to a method for the determination of core parameters using a gamma-gamma logging technique.

In this field of application core samples are taken from formation layers to be used for experiments and investigations in order to characterize said formation layers and to determine parameters of said samples as desired.

In particular the present invention relates to a method for the determination of density, diameter lithology parameter and elemental concentration of at least one core sample arranged in a detecting device, said method comprising detection of gamma rays arising from radiation induced by an external gamma ray source, optionally cesium (Cs), irradiating said sample.

It is known from US patent specification No. 4 529 877 to apply a gamma-gamma density logging technique in order to determine formation densities within a borehole.

Said known technique uses a Cs source arranged in a logging tool and irradiating the surrounding formation further provided with two differently longitudinally spaced detectors, i.e. a short spaced (ss) detector and a long spaced (ls) detector, with which gamma ray spectra are observed. By using both detectors mud cake and/or casing contributions are eliminated.

Furthermore the detection ranges of each detector are subdivided in four energy windows, three of which covering each a main energy peak of a natural gamma-ray source, usually of thorium (2.61 MeV), uranium (1.70 MeV), and potassium (1.46 MeV), and the fourth of which detecting scatter radiation from the Cs-source.

Within said three windows natural source count-rates are measured allowing for both detectors the correction of scatter radiation count-rates from the fourth window with respect to naturally occurring gamma rays by using a conventional so called stripping technique.

Finally the ss and ls count-rates corrected result in the determination of formation density values, taking in account Compton scattering conditions for the said Cs rays. In addition, concentrations of natural source elements can be computed simultaneously.

The above known method clearly relates to a logging technique within a borehole. Consequently depth correlation for the data as acquired and processed has been obtained inherently.

However, it should be necessary to make depth correlation fits between the above said data being wireline log data and core sample data. Besides it could be desirable to obtain further detailed information from core samples.

Furthermore it has appeared that the above data fitting is hampered by the conditions of said core samples as taken, i.e. the cores are broken, parts of cores have been lost resulting in irregular cross sections, different parts of cores have been rotated in different ways and over different angles, and lengths of cores are changed.

Thus it is an object of the invention to get a reliable depth correlation fit for core samples using a gamma-gamma logging technique.

Further it is an object of the invention to accomplish a high level accuracy in the gamma-gamma logging technique used in obtaining core sample information.

It is another object of the invention to get both information about a number of core parameters and extensive parameter data of said samples.

The invention therefore provides a method for the determination of density, diameter, lithology parameter and elemental concentrations of at least one core sample arranged in a detecting device, said method comprising the steps of detecting gamma rays arising from both natural sources within said sample and from radiation induced by an external gamma ray source, optionally cesium (Cs), irradiating said sample, - applying a gamma-beam-width greater than the sample diameter, - using a gamma ray detector provided with several count-rate windows over two main energy regions, - detecting first gamma rays in a first low energy region including a first set of count-rate windows and related to radiation induced by the gamma source thereby obtaining first count-rates, - detecting second gamma rays in a second high energy region including a second set of count-rate windows and related to said natural gamma ray sources, thereby obtaining second count-rates, - defining by means of said first count-rates indicators related to density, diameter and lithology parameter of said sample, and - processing said second count-rates to determine concentrations of said natural sources.

Advantageously said first and second set of windows cover substantially all energy ranges of relevant physical effects to be awaited, said indicators coupling count-rates as measured in the respective windows with core sample physics thereby modelling sample features closely. Thus reliable core sample parameter values are obtained.

Furthermore it is advantageous that said core sample is sampled from a core length which is supplied controllably to said detecting device. Preferably said core length is supplied continuously by means of a conveyor belt.

Said detecting and measuring set up allows core characterizing and depth correlation in a simple and reproducible way.

A further advantageous attainment of the invention involves a method for the correction of said first and second count-rates for effects of background radiation, secondary radiation, and variation in sample diameter and density, correcting said second count-rates by means of, - a correction procedure for said background radiation, and - a correction procedure for multiple incidence Cs events, correcting said first count-rates by means of a correction procedure for secondary gamma ray contributions from said natural sources, and correcting both first and second count-rates by means of a correction procedure for normalizing said count-rates to constant core diameter and density.

Also close physical modelling results in advantageous correction procedures for the said effects. As a result accurate parameter data of said core samples are obtained.

Furthermore the invention provides a method for the correction of density, diameter, lithology parameter and elemental concentrations of at least one core sample arranged in a detecting device for the detection of gamma rays arising from natural sources within said sample, said method comprising, - using a gamma ray detector provided with several count-rate windows over two main energy regions, - detecting first gamma rays in a first low energy region including a first set of count-rate windows, - detecting second gamma rays in a second high energy region including a second set of count-rate windows and related to said natural gamma ray sources, thereby obtaining second count-rates, - processing said second count-rates to determine concentrations of said natural sources, - correcting said first count-rates by means of a correction procedure for secondary gamma ray contributions from said natural sources, - correcting said second count-rates by means of a correction procedure for said background radiation, and - correcting both first and second count-rates by means of a correction procedure for normalizing said count-rates.

The above correction method has the advantage of determining and correcting parameters, for example diameter and density, of core samples containing natural gamma ray sources by measuring only natural gamma radiation. Thus an independent correction procedure for core sample parameters is obtained.

The invention will now be described in more detail by way of example with reference to the accompanying drawings, in which: - fig. 1 is a diagram showing the relation between an indicator Tc, which will be explained hereinafter, and a first density/ diameter-couple for core samples, taken from a set of calibration bars, which relation has been obtained in accordance with the present invention; - fig. 2 is a diagram showing the relation between indicator Ts, which will be explained also hereinafter, and a second density/ diameter couple; - fig. 3 is a diagram showing the interrelation between the above said indicators T and T , represented respectively in fig. 1 and c 5 fig. 2;; - fig. 4A and fig. 4B are diagrams showing the relation between the pe-value which will be explained hereinafter and respectively indicators T 1 and T 2 corresponding to different pe-value-ranges pl p2 for different diameter ranges; and - fig. 5 shows a schematic drawing of the equipment used for carrying out the method in accordance with the invention.

When cores are taken from formation layers for petrophysical and geological purposes said cores are tested on a routine basis for natural gamma radiation. However, gathering and transporting core lengths result in extra handling of said cores and often result in misplaced cores. It is estimated that for more than 50% of all said cores some parts thereof are misplaced before any core sample is selected.

Curves depicting total gamma radiation, density and/or pe-value as functions of core depth may be used to facilitate depth correlation between core and wireline log data. Wireline depths and field recorded core depths seldom agree. Coring depth is measured e.g. from tape-measuring the drill pipe, while the logging depth is obtained from the logging cable. Differences of the order of 1 m are common. Furthermore at the surface the core is usually broken into pieces to fit in boxes resulting in gaps and misplaced core.

Therefore reliable core gamma measurements require that the core be aligned properly allowing the results to be compared to the data from wireline logs. Moreover, it is desirable that the alignment of the core can be done while the instrument is running during said operation spectral analysis of the natural gamma radiation being allowed.

It will be clear to those skilled in the art that a sufficient service not only involves checking the distribution and levels of natural gamma radiation but also requires correction and measurement procedures directed to a core characterization method which covers parameter ranges as wide as possible. Consequently investigations have been made how to improve parameter resolution, core characterization and depth correlation.

Said investigations have revealed that information about density and diameter is required to correct natural gamma radiation levels. Maybe of the same importance density, diameter and pe-curves are useful tools over "clean" intervals where the total gamma ray flux is low or featureless. This may happen for clean sandstones with variable porosity. Another interesting case is that of the clean carbonate reservoir with porosity associated with dolomitization. Here both gamma and density variations may be minor, but pe-variation substantial.

Before disclosing a method for the determination of density, diameter, lithology parameter and elemental concentrations of such a core, in particular of core samples supplied to a detecting device in a controllable way, and moreover a correction procedure for the different gamma ray contributions as presented in this invention a short summary of some gamma log physics has to be given.

In nature, several long-lived radioactive nuclides occur. In subsurface rocks those of potassium (K), uranium (U), and thorium (Th) are responsible for virtually all of the measurable gamma radiation.

The radioactive disintegration of U and Th occurs in a sequence of steps. Each step involves emission of radiation, in particular of gamma radiation of various characteristic energies.

After every step, a new radioactive isotope results until the stable lead isotope is reached. Each isotope disintegrates with a certain half-life. The corresponding half lives may be very long to 1010 (up to 101 years). The daughter isotopes must be in equilibrium with their parent isotopes to obtain stable gamma radiation 238 235 spectra. For uranium, both 238U U and U will contribute to the radiation and must be present in constant proportion (usually 108 235 of the gamma photons are emitted from the isotopes from the 235U series). The main energy peaks of natural U (the 238-isotope) and Th gamma radiation are respectively 1.70 MeV and 2.61 MeV.

For the K-source the half life is in order of 10 years and the corresponding main energy peak is 1.46 MeV.

137 In gamma-gamma logging techniques conventionally a Cs gamma ray source is used having a half life of about 30 years and a main energy peak of 0.66 MeV.

In general the elemental abundances for said natural sources U, Th, and K in oil-well cores are up to respectively 30 ppm, 30 ppm and 8% (m/m). For external Cs gamma ray sources radiation levels can be chosen. However, safety requirements have to be regarded. In the set up, which will be shown in detail hereafter a 10 pCi - 137cs source is used. Then a regularly shielded source will have a radiation level which is negligible compared with the background level.

The different contributions to the gamma ray spectrum which will be detected and which has to be analyzed have to be discussed now.

Only two modes in which gamma rays interact with matter are relevant for the present invention.

The first type of interaction is Compton scattering, in which a gamma photon interacts with an electron. In this type of interaction the binding energy of the electron is negligible and the electron essentially behaves as a free particle. In such an interaction an incident photon will transfer only part of its energy to the electron and the remainder of the energy appears as a secondary, scattered photon of lower energy.

The second type of interaction is the photoelectric effect (in the following designated as pe and used as an index), in which the gamma photon interacts with an electron that is not free, but bound to an atom. Here the photon is fully absorbed, the electron is ejected, and the remaining atom is ionized.

Both types of interactions will attenuate the intensity of the original gamma radiation. For each interaction an attenuation coefficient, , can be defined as the relative decrease in intensity (dI) per unit length (dx) of a linear beam of radiation: 1 di - I dx (1) with I as intensity of the gamma rays originating from a source, whereas p (in units m 1) is usually referred to as the linear attenuation coefficient. The linear attenuation coefficient is a function of the electron density of the material through which the incident beam passes.A mass attenuation coefficient is also defined by normalizing the linear attenuation coefficient to the bulk density: U - H/P (2) For Compton scattering the mass attenuation coefficient is virtually independent of the atomic number (Z) and only slightly dependent on the energy of the photons (E) (in units eV): V -Z0 E -0.4 (3) S with index s for scatter For the photoelectric effect the mass attenuation coefficient increases strongly with atomic number and decreases strongly with the energy of the photons: v =3 .6 (4) -3.2 (4) pe The actual exponents in the equations (3) and (4) do vary slightly with energy and atomic number. The above values are valid for lower energy and atomic numbers between 10 and 20.

Up to now for the determination of elemental concentrations of e.g. U, Th, and K the primary radiations were reduced to only a single main energy peak for each element, i.e. the above said 1.70 MeV, 2.61 MeV, and 1.46 MeV spectral lines. The Th line with the highest energy will create a spectrum at lower energies and increase the count-rates in the vicinity of the U and K lines. The U line will increase the count-rates near the K line. However, the increases are proportional to the amount of the radiation at the higher energies, because for a given detection system the shape of the spectrum is constant.

So from the count-rate at the Th line it is possible to calculate the lower-energy contribution from Th and subtract it from the count-rates at the K and U lines. In the same way the extra contribution from U to the count-rate at the K line can be eliminated. This in principle is the spectrum stripping technique.

In the real situation where U and Th have many primary lines, the relative shape of the spectrum still remains constant and the same spectrum stripping technique can be applied, except that the Th line must be corrected for contributions from small subsidiary U peaks.

In principle only three energy windows are required to calculate elemental contents of K, U, and Th. In practice low count-rates result in unacceptably large statistical variance in the estimates. Experiments have shown that use of more than three windows will significantly improve the measurement by reducing statistical variations.

The measurement of the density (p) uses a range of gamma photon energies where Compton scattering predominates over other processes. For a narrow beam of gamma rays the intensity of the primary photons (I ) can be found using equations (1) and (2) over the core diameter (d): ln (I ) - vpd (5) When the energy is known the value for v is known. When the core diameter is known the density may be calculated. When a beam wider than the core is used, a different relation exists between the scattered radiation intensity Is, , density and diameter (p, d). This allows the independent calculation of density and diameter values.

Furthermore it is possible, using low energy gamma photons, to determine the photoelectric mass attenuation coefficient (vpe) by a pe separate measurement. In wireline logging a pe-value estimate, written as pe - (Z/10) 3.6 (replacing vie), is used. On the pe contrary, the approach in accordance with the present invention allows a direct determination of pe-values for Z < 20.

Thus, the main aspects from which the present invention starts, involves the division of the gamma ray spectrum in two regions. The high energy (0.75 to 3.0 MeV) region of the spectrum consists of primary peaks due to natural radioactive core constituents, K, U, and Th. Radiation (primary and scattered) from the Cs source does not contaminate this portion of the spectrum except in very minor ways (multiple-incidence occurrences). The intensities of these high energy primary peaks when suitably corrected are used to compute the elemental concentrations of K, U, and Th. The low energy region of the spectrum (0.04 to 0.75 MeV) consists of many components including the primary Cs peak, tails from higher-energy primary core constituent peaks and radiation arising from gamma ray scattering in the core and detector crystals.The intensity of the scattered components from primary Cs radiation varies in relatively predictable ways with the mass and size of the core, thus allowing a determination of density and diameter. The very lowest energy window is used to determine pe-value.

Furthermore it has appeared advantageous that within said two energy regions twelve energy "windows" have been identified and implemented. In each window the aggregate number of counts resulting from gamma photons impinging on a detector device unit is used for further defining and processing purposes in accordance with the invention. However, dependent on particular conditions or circumstances, other choices for the number of windows could be preferred.

Within said twelve windows the first four windows fall into the low-energy region (0.04-0.75 MeV) and are used to determine density, diameter and pe-values, by recording the count-rate data induced by a radioactive 7Cs source. The last eight windows fall into the high-energy region (0.75-3.0 MeV) and are used to determine the total gamma radiation and the individual elemental contents of K, U, and Th. Further details of the selection of the above windows as presently used, will be discussed hereafter. An overview of said features is given in the corresponding table I.

TABLE I Window Energy (keV) Comments 1 40- 85 Photoelectric Effect 2 275- 395 Low Compton Scattering 3 400- 520 . High Compton Scattering 4 575- 720 Cesium Peak 5 770-2835 Total GR Window 6 840-1045 Low Energy U, Th Peaks 7 1050-1320 Low Energy U Peaks 8 1340-1570 K Peak 9 1660-1890 Main U Peak 10 1970-2400 High Energy U Peaks 11 2405-2830 Main Th Peak 12 2835-3060 Tail of Spectrum Windows 1 to 4 are used for the determination of the density, the diameter and the pe-value. Window 5 is used to calculate the total gamma radiation and windows 6 to 11 are used to calculate the individual contents of K, U, and Th. Window 12 monitors the cosmic background variations.

As shown in detail in the above table window 4 is selected 137 around the primary energy of the Cs radiation of 0.662 MeV. The 137 energy range is 0.575 to 0.720 MeV. Cs decays by ss emission, to 137 give an ion of Ba which attracts an electron. The subsequent cascade of an electron into the Ba K-shell generates so-called Bremsstrahlung with a maximum energy of 0.039 MeV. This Bremsstrahlung creates a peak in the very low end of the spectrum, where the photoelectric effect dominates. To avoid it the first window is set just above the edge of this peak and runs from 0.040 MeV to 0.085 MeV. The high end of window 3 is picked to lie just above the Compton edge in the continuous part of the spectrum of the 137 radiation, which is at about 0.480 MeV.The low end of window 3 is chosen at about 0.4 MeV, a point below which there is 137 an increase in slope in the primary Cs spectrum.

The low end of window 2 is selected at 0.275 MeV where the photoelectric attenuation is less than 10% of the total attenuation for calcium to avoid lithology effects in the density calculations.

The high end of window 2 is adjacent to the low end of window 3.

Thus window 2 extends from 0.275 to 0.395 MeV and window 3 from 0.40 to 0.52 MeV.

Windows 8, 9 and 11 were selected to extend over the main peaks of K, U, and Th respectively. Window 10 fills the gap between the U and Th windows. Only the boundaries of window 10 can be selected with some degree of freedom. The low end of window 6 is chosen at 0.840 MeV at the start of a discernible U peak from the U-line spectrum and ends at 1.05 MeV, just beyond a lower energy Th peak. The low end of window 7 begins at 1.05 MeV and ends at 1.32 MeV and extends over two recognizable U peaks and no Th peaks. Many windows do not actually touch as can be seen from Table I.

Extending the windows further only introduces extra backscattered radiation from higher energy peaks, but does not decrease the statistical variations in the results. Indeed a marginal increase was observed in the statistical errors when the windows were expanded to touch each other.

Corresponding to the above division in two main energy regions and further selection of windows in the first low energy region covering the first set of count-rate windows first count-rates will be obtained, and in the second high energy region covering the second set of count-rate windows second count-rates will be obtained.

Usually for density measurements a narrow beam of gamma radiation is used. The attenuation of the intensity of this beam is an exponential function of the product of the density (p) and diameter (d) which can be presented by combining the formulas (1) and (2) as I/Io - exp (-vpd) (6) in which v is the total mass attenuation constant, which is usually quite independent of the composition and density of the material investigated. However, in order to apply the appropriate corrections to the gamma radiation measurements on cores it is necessary to have separate estimates of the diameter value and density value.

A wide beam was selected which cuts through a slice of core and is always wider than the core diameter. It can be shown that the attenuation of the primary radiation is a function of the 2 parameter couple (pd ) whereas the intensity of the scatter radiation is differently dependent on the density and diameter. For 137Cs radiation the only important mechanism for attenuation is Compton scattering. This means that the attenuation of the primary radiation yields an amount of scattered radiation equal to the amount lost from the primary radiation. The scattered radiation is deflected over all three dimensions in the core and will be attenuated differently than the primary radiation, allowing the effects of density and diameter to be evaluated separately.

The density and the diameter of the core are estimated from the count-rates in the first four windows introduced by the radiation from the Cs source. In the particular algorithm as disclosed hereafter only windows 3 and 4 are used. The effects of background and natural gamma radiation will be ignored in the following reasoning, because they can be adequately corrected for during the calculation.

When no core is present in the detecting device, count-rates recorded in the first four windows arise from the primary photons emitted by the Cs-source. In the lower energy windows (1, 2, and 3 as shown above) count-rates result from primary gamma rays which deposit only a part of their energy in the detector. If the number of primary counts per second varies, the number of counts in all four windows will vary in proportion. The number of primary counts will decrease when a core is present because of the interactions between the gamma rays and the core.

In order to determine the density and diameter as sketched above indicators are defined which are composed of count-rates I detected in the above said windows.

The count-rates in the present windows 3 and 4 as detected in the core free detecting device (hereinafter indicated with an index f for free), respectively Compton scatter window and indicated with above said index s for scatter, and Cs-peak window with an index c for Cs and always in first position, are indicated as Isf and Icf.

On the contrary when-said device contains (for which an index c of contain, always in second position, will be used) a core sample the count-rate in window 4 is due curly to the primary photons because the photons that are scattered in the core will have a lower energy. The count-rate in window 3 (or scatter window s) is due partly to primary gamma rays and due partly to secondary gamma rays which result from scattering in the core. The primary radiation part in windows 3 and 4 equals respectively IscIcc/Icf and I . For cc window 3, the difference between the total count-rate and the primary radiation part is due to the interactions occurring in the core.For window 4 (or Cs-window c), the difference between the count-rate without a core and the count-rate with a core present in the cavity is due to the attenuation in the core. For both windows, it is necessary to normalize to the situation without a core to compensate for small changes in source strength and other effects.

For that purpose two indicators are defined: T - = (ICf-Icc)/Icf c cf cc cf (7) and

with indices first c, second c, s and f, as mentioned above. It will be clear that said indicators are dimensionless. When taking in account general geometrical conditions of the detecting device and the above said wide beam set up the following expressions for T and T can be derived: e s -c Tc/51n (pd2) - k (9) c and Ts/ln (d2/p) k k (10) S with kc and k5 being constants.

It will be clear from the above expressions (9) and (10) that by detecting count-rates in the said relevant windows and consequently by combining (9) and (10) the density p and diameter d can be determined, i.e. respectively by dividing and multiplying (9) and (10).

Referring now to fig. 1, the vertical axis represents parameter couple dVp in units cm Mg /2 3/2, whereas the horizontal axis represents the indicator T as defined in the above in c arbitrary units. The horizontal axis has a linear scale, whereas the vertical axis has a logarithmic scale.

In fig. 2 the vertical axis represents parameter couple d/Xp in units cm Mg -1/2 m 3/2, whereas the horizontal axis represents the indicator T as defined in the above in arbitrary units. The S horizontal axis has a linear scale, whereas the vertical axis has a logarithmic scale. The symbols used in this graph are density 3 values in units Mg/m Both graphs result from detections and processing of count rates for T and T on core samples from cylindrical calibration c s bars. The expressions (9) and (10) can be checked allowing the determination of the relevant constants.

From fig. 3, with T on the vertical axis and T on the S c horizontal axis, it can be seen that the diameter is better resolved than the density. The measurements shown in this fig. have been made for calibration bars composed of eight different materials for which corresponding eight different symbols are used.

For T > 0.52 the values for the density and diameter are obtained c by intrapolation as shown by the boxed area.

With the detecting device as designed to determine density and diameter values it is fairly straightforward to calculate the above mentioned pe-value. The only extra feature needed is a low energy window as explained above. Thus window 1 from table I will be used as pe-window, indicated with an index p for photo electric, detecting count-rates Ipf and I (with again f for core free and c pc for core containing).

As explained above the pe-value is very sensitive to changes in the average atomic number and provides another tool to depthcorrelate core and log data. Again an indicator is defined.

However, for pe-values with pe < 10 and for pe-values with pe > 10 two different indicators, respectively Tpl and T should be used pl p2 advantageously. The following expressions are defined:

In a similar way as for T and T relations can be derived which S couple the above Tpl and Tp2 to pe-values. The general form of such a relation is pe - a T -b (13) i I with i - 1 or 2, respectively for pe-values < 10 or > 10, and a.

and b. as constants related to the diameter d.

Referring now to figs. 4A and 4B the vertical axes represent pe values in arbitrary units whereas the horizontal axes represent respectively the indicators T and T . Horizontal and vertical pl p2 axes in both figs. have linear scales.

Both graphs result from detections and processing of count-rates for T 1 and T on core samples from cylindrical pl p2 calibration bars. The different symbols in said graphs are indicating different diameter values for the calibration bars used the values ranging from 2 inch to 5 inch. The relation (13) can be checked and the different application ranges for different pe-values can be determined. Furthermore it has to be noted that pe-values over 10 are uncommon and may very well result from spurious count-rates or contaminated core as a result of the above pe can be determined when Tpl has been measured in well known pl conditions for a. and bi.

1 1 In accordance with the invention for the determination of the concentrations of said natural sources an improved method is used.

As explained above in the high energy part of the spectrum all the detected radiation comes from either K, U, or Th after the background radiation is eliminated. Thus the contents of the three elements can be calculated using three windows which are appropriately selected in the spectrum because the count-rates in the three windows yield three equations with three unknowns. When spectral analysis of natural gamma radiation started, this method was referred to as the "spectrum s,tripping".

In the present design there are eight windows available in the high energy part of the spectrum and from any combination of three windows out of these eight one can calculate a value for the contents of K, U, and Th. In the absence of statistical fluctuations and calibration errors any combination would yield the correct answer for all three elements. However, in reality these fluctuations are not negligible. A natural idea for reducing the error is to average over some combinations of three windows. If more than three windows are used to calculate the contents the error should become smaller.

Usually when more than three windows are used, the coefficients for the equations are determined using least squares linear regression. This technique was attempted but the results were unsatisfactory. It now appears that linear regression may not be applied to this problem, since it assumes that the independent variables are mutually (statistically) independent. However, for more than three windows all extra windows are completely dependent on the first three. Furthermore, the errors in the count-rates are to some extent correlated to the errors in the contents because these are not accurately known. For the longer counting times used during calibration this becomes important and represents another violation of the conditions for validity of linear regression.

Another method is designed to avoid the short comings of the linear regression technique. This method is called "Monte Carlo" method mainly because a random number generator is used to introduce well-known errors, which exactly obey a Poisson distribution. Monte Carlo methods as such are known to those skilled in the art and will not be explained in detail.

With the aid of the relations for the count-rates in the various windows and the known background levels the total raw count-rates for different amounts of the various elements were calculated. To these count-rates Poisson-distributed random fluctuations were added using a pseudorandom number generator, with the mean value equal to the product of the count-rate and the counting time. The advantage of this procedure is that the errors in the observed count-rates are not correlated with the errors in the contents of the various elements. Also the expected variance is exactly equal to the total number of counts, so that the conditions for weighted linear regression are not violated.

Next sets of windows are selected and consequently sets of equations for said windows comprising coefficients for the different window contributions which both obey the above stated requirements and allow a modified regression technique when solving said set of equations. Additionally a calibration procedure to calibrate the different possible sets is used.

Results of the processing procedure as disclosed are shown in Table II.

TABLE II VARIABLE ERROR RANGE K 0.25 1-58 U 1.0 2-10 ppm Th 2.0 2-20 ppm GR 6.0 40-250 API The above range is the range of values over which the contents were varied in the calibration bars (API-units for gamma logs are known to those skilled in the art and will not be defined here). It is felt that the accuracy is sufficient and corresponds to that for K, U, and Th values obtained from wireline data. The accuracy of the total gamma ray is better than the accuracy of the corresponding curve obtained thus far by using conventional methods.

As disclosed above complex problems have been solved in obtaining density, diameter and pe-values and in processing elemental concentration values.

However problems arise when trying to make corrections on the aforesaid values caused by background, density, and diameter effects.

For the high energy part of the spectrum corrections have to be made for background radiation, and for the density and diameter of the core. Furthermore, said background radiation itself is also affected by the density and the diameter. Additionally in the range between 0.7 and 1.4 MeV corrections must be made for multiple incidence events, itself also dependent on the density and the diameter of the core.

For the low energy part of the spectrum, corrections are required for the background radiation and the contributions from the natural gamma radiation. The background radiation in this energy range is found to be independent of the density and diameter of the core. The contribution from natural gamma radiation can be evaluated from the high energy part of the spectrum.

The effects of the variation of the background radiation as a function of time have been significantly reduced by the lead shielding. The background radiation levels as seen by the detectors are now low with respect to the normal gamma radiation levels.

Variations of a few per cent or less with time are thus inconsequential.

However, although the background levels are low in the present detecting device, and are fairly constant (within 2%) they do vary with the density and the diameter of core. For the higher energy windows typical attenuation of background radiation due to the core is between 5% and 10%.

Because background radiation is subject to the same attenuation processes as all radiation the effect of the density 2 and the diameter of the core are involved resulting in a pd2 term in a correction factor.

For the windows 5 to 11 the following correction factor on 1bif (with index b for background, f for free and i - 5 ...., 11): 1bic bif (l-Apd2) (14) with A a constant obtained from empirical data.

For windows 1 to 4 no effect is noticeable. Also for these windows the background correction is significantly less important, since the induced gamma radiation results in count-rates much higher than those due to natural radiation.

It has been found that the remaining error after correction for the background radiation cc,;esponds to about 0.03% K, 0.3 ppm U, and 0.6 ppm Th.

A second source of errors which affects count-rates is the effect of multiple incidence radiation.

When a source is present some radiation was detected in the higher energy windows 5, 6 and 7. No extra radiation in the other high energy windows was observed. Two Cs sources of different strengths (10 p Ci and 1 p Ci) were used. The results indicated that the extra radiation was strictly proportional to the source strength. Because the extra radiation seemed to be limited to energies below 1350 keV, it is assumed that the detectors may be seeing two gamma rays which strike any of the three detectors simultaneously before the deadtime circuitry is activated. The energies of the two photon interactions are then added and because the Cs energy is 660 keV the maximum energy detected in double incidence occurrence is 1320 keV. Furthermore, the chances of triple and higher incidence occurrences are very much lower and need not be considered.

The double incidence radiation could easily contribute 20 API units to the total gamma ray curve and obviously must be corrected for. It is also attenuated by the core. For all three windows, the 2 same relationship with pd2 holds. The relationship is found to be rather complex and can be given as a correction function Br, Mimic - Imif Br (pd2) (15) with index m for multiple incidence, f for free, window number i - 5, 6, 7, index r indicating a mass range dependency, and the function B as a power series in pd2 with different coefficients for different mass ranges to be determined empirically.

This correction will bring the effect of multiple incidence radiation down to the level of 1 API unit on the total gamma ray curve and for K, U, and Th the effect is negligible.

The count-rates taken over the whole energy spectrum after background and multiple incidence corrections in all windows vary with the amount of core between the detectors if everything else is kept constant. When the core is broken its shape is rather irregular. However, the method of calibration always assumes the core is cylindrical and can be characterized by an effective diameter. When the weight percentage of the components is kept constant, the amount of radioactive'nuclei varies in proportion to the density and to the cross-section. However, when the diameter increases the radiation is attenuated more by the extra matrix material present.

This effect can be described by a first correction function C1, applied on a proportionality constant I. for window i: C1 Cl(p,d) - d exp(-vipd) (16), with vi as the effective attenuation coefficient for said window i, thereby obtaining corrected count-rates Idic which represent corrected diameter count-rates in window i for the core containing detecting device, respectively indicated by the other two indices d and c.

Tests have been carried out on calibration bars having diameters of 5, 7.5, 10 and 12.5 cm.

It was expected that the value for the effective attenuation coefficient would vary with energy of the corresponding window more or less in correspondence with the actual attenuation coefficient.

This was not the case. The best results were obtained with a value of 0.060 for all windows. This value is a weighted average, in which the weighting factor was selected to be proportional to the square root of the count-rate. The dispersion in the results is due to the variations in the contents of radioactive elements, variations in the density and diameter and also due to the different count-rate levels. When we plot the ratio of net count-rates large fluctuations can be expected at low count-rate levels.

If in the reference conditions the density is pO and the diameter is d then all count-rates can be normalized to these o reference conditions using the above equation (16) multiplied with a second correction function C2 for said window i applied on Idic'

with s is a constant, thereby obtaining normalized (n) diameter count-rates Idin Normalizing within 20% has been proved.

For the calculations of the density, diameter and pe-value of the core only the counts induced by the Cs source are relevant. In particular, background radiation and natural gamma radiation must be eliminated from the total 'count-rates in these windows. However, for the windows 1 to 4 the background radiation is relatively small and virtually independent of the density and diameter of the core, while the natural gamma radiation is dependent on the density and diameter and on the contents of K, U, and Th in the core.

A correction procedure which was proven reliable employes the relevant peak window related to high energy count-rates originating from natural gamma ray sources. Next the low energy count-rates from a Cs free detection device are arranged in a set of equations which are then processed by means of the Monte Carlo modified regression algorithm as disclosed above. The resulting low energy count-rates are compared with said first count-rates yielding a final correction for said first count-rates. This final correction results in improved results up to 5%.

It is noted that for proper corrections in the high energy part of the spectrum the data from the low energy part is needed, which should be first corrected using the high energy part. The same holds true, mutatis mutandi, for the low energy part of the spectrum. Because the corrections are small an iterative procedure will converge very quickly. Presently a fixed three-step iteration procedure is applied in the calculation program.

It will be clear for those skilled in the art that also a method for the correction of parameters, for example diameter and density, of core samples containing natural gamma ray sources by measuring only natural gamma radiation is possible. In that case no use is made of Cs-induced count-rates and thus an independent correction procedure for core samples is obtained.

However when the above disclosed correction steps should be made some restrictions have to be observed.

Although count-rates can be determined even in the low energy windows their amounts will be low. Thus much longer counting-times could be necessary. Furthermore correcting multiple incidence Cs-events can be avoided properly since the chance of such events is substantially zero.

Referring now to fig. 5 a schematic drawing illustrates the equipment as presently used for carrying out the above disclosed methods for both the determination of several core parameters and correcting the parameter values determined. In the set up as shown a detector device 1 comprises advantageously three scintillation detectors made of sodium iodide (NaI) doped with thallium (T1) and connected with a photo multiplier tube 2 which is coupled to a pre-amplifier 3 and an amplifier 4 and to a high voltage supply 5.

A multi channel analyser (MCA) 6 to be considered as the most important electronic component of the system, comprises a number of channels which are divided over the above mentioned windows. In the set up as chosen the MCA has 2048 available channels which are split in two equal groups of 1024. At regular time intervals the two groups are swapped. One group is busy recording the counts in the various channels, while the other group is being used to write the previously obtained values to a computer 7 which controls nearly the entire procedure during normal operation. On a given signal the counting is stopped. The group that was read into the computer is cleared and the two groups swap their functions, so that the recording will proceed into the cleared group. The computer also monitors the (depth and) speed of the core.

A lead shielding 8, further comprising a copper rack 9, reduces the amount of environmental radiation able to reach the detectors. The copper rack is a further shielding against X-rays generated in the lead thus forming an error source for the count-rates detected. A steel housing 10 further enhances the shield effect.

Through an aluminium liner 11 a transporting device loaded with cores conveys core lengths during which transporting procedure a Cs source within a source holder 12 irradiates a core sample. In the current design the core is moved in troughs made from halved plastic pipes on a three-meter conveyor belt 15 in front of the instrument; behind the instrument the core moves over rollers. The trough sections are about 3 meters in length. However this systems does not guarantee that slippage will not occur. The gamma ray intensities must be recorded as a function of depth rather than time. This requires a depth-measuring device 16 to record the actual depth of the core at every point in time. In the embodiment as presented said device consists of a roller that is rotated by the trough holding the core.On the axle of this roller a chopper is mounted which interrupts a light beam from a small light-emitting diode at exactly every centimetre displacement.

Counting these interruptions gives a relative depth record of the core. An electronic system is added which counts the total number of interruptions which will be taken in account when indication said depths.

Some additional features from fig. 5 are mentioned yet.

Reference number 17 represents a cable conduit throughout the shielding 8. Furthermore a roller 18 serves as a conduit for the core lengths supplied on the belt from left to right into the detecting device, the belt rotating clockwise. Respective arrows presented in fig. 5 are illustrating said movements.

Various modifications of the present invention will become apparent to those skilled in the art from the foregoing description and accompanying drawings. Such modifications are intended to fall within the scope of the appended claims.

Claims (21)

1. A method for the determination of density, diameter, lithology parameter and elemental concentrations of at least one core sample arranged in a detecting device, said method comprising the steps of detecting gamma rays arising from both natural sources within said sample and from radiation induced by an external gamma ray source, optionally cesium (Cs), irradiating said sample, - applying a gamma-beam-width greater than the sample diameter, - using a gamma ray detector provided with several count-rate windows over two main energy regions, - detecting first gamma rays in a first low energy region including a first set of count-rate windows and related to radiation induced by the gamma source thereby obtaining first count-rates, - detecting second gamma rays in a second high energy region including a second set of count-rate windows and related to said natural gamma ray sources, thereby obtaining second count-rates, - defining by means of said first count-rates indicators related to density, diameter and lithology parameter of said sample, and - processing said second count-rates to determine concentrations of said natural sources.

2. The method as claimed in claim 1, wherein the first set of count-rate windows comprises two windows with which an independent determination of sample density p and diameter d is obtained.

3. The method as claimed in claim 2, wherein for the said two windows two corresponding indicators T and T are defined C S comprising count-rate ratios of count-rates from the corecontaining and core-free detecting device, respectively for the window containing the source peak energy and the window containing Compton scattering energy.

4. The method as claimed in claim 3, wherein for said indicator 2 T a relation, Tc/ln(pd ) - k , is used (with k being a constant).

c c c c

5. The method as claimed in claim 3, wherein for said indicator 2 T a relation, Ts/ln(d2/p) - k5, is used (with k being a S S constant).

6. The method as claimed in claim 4 and 5, wherein said independent determination of sample density and diameter is obtained by combining the relations for T and T c s

7. The method as claimed in claim 1, wherein the first set of count-rate windows comprises at least one window with which a pe-value representing gamma ray attenuation caused by the photo-electric effect is obtained.

8. The method as claimed in the claims 1 and 3, wherein at least one indicator T is defined comprising count-rates obtained in the p said windows from the core-containing and core-free device.

9. The method as claimed in claim 7 and 8, wherein for said pe-value and indicator T a relation, pe - aT + b, is used (a and p p b being constants).

10. The method as claimed in claim 1, wherein the second set of count-rate windows comprises at least windows which each cover a main energy peak of each respective natural source, within said windows respective count-rates being obtained, said count-rates being arranged in a set of equations which are processed by means of a regression algorithm calibrated by using a Monte Carlo method.

11. The method as claimed in claim 1, wherein the core sample is sampled from a core length controllably supplied to said detecting device.

12. The method as claimed in claim 11, wherein said core length is supplied continuously.

13. The method as claimed in claim 12, wherein said core length is supplied by means of a conveyor belt.

14. A method for the correction of density, diameter, lithology parameter and elemental concentrations of at least one core sample arranged in a detecting device for the detection of gamma rays arising from natural sources within said sample, said method comprising, - using a gamma ray detector provided with several count-rate windows over two main energy regions, - detecting first gamma rays in a first low energy region including a first set of count-rate windows, - detecting second gamma rays in a second high energy region including a second set of count-rate windows and related to said natural gamma ray sources, thereby obtaining second count-rates, - processing said second count-rates to determine concentrations of said natural sources, - correcting said first count-rates by means of a correction procedure for secondary gamma ray contributions from said natural sources, - correcting said second count-rates by means of a correction procedure for said background radiation, and - correcting both first and second count-rates by means of a correction procedure for normalizing said count-rates.

15. The method as claimed in claim 1, further comprising the correction of said first and second count-rates for effects of background radiation, secondary radiation, and variation in sample diameter and density, correcting said second count-rates by means of, - a correction procedure for said background radiation, and - a correction procedure for multiple incidence gamma source events, correcting said first count-rates by means of a correction procedure for secondary gamma ray contributions from said natural sources, and correcting both first and second count-rates by means of a correction procedure for normalizing said count-rates.

16. The method as claimed in claim 14 or 15, wherein the correction procedure for said background radiation applied to said second count-rates employes a correction factor in a window (i) on background (b) count-rate 1bif detected in the core free detecting device, l-Apd (A being a constant), thereby obtaining Ibic representing the corrected background count-rate for the corecontaining detecting device.

17. The method as claimed in claim 15, wherein the correction procedure for multiple incidence gamma source events for said second count-rates employes a correction function in a window (i) on multiple incidence (m) count-rate Imif detected in the core free (pd2 for mif detecting device, B - B Br(pd ) for mass range r, thereby obtaining r r Mimic representing the corrected multiple incidence count-rate for the core containing detecting device.

18. The method as claimed in claim 14 or 15, wherein the correction procedure for normalizing said first and second count-rates employes a first correction function in a window (i) applied on a proportionality constant Ii: C1 -C - pd2 1 1 1(p,d) exp(- u.pd) (Vi is the effective attenuation coefficient for said window (i)), thereby obtaining Idic representing the corrected diameter (d) count-rate for the core-containing detecting device, and a second correction function for said window applied on 1disc'

(K being a constant, and p0 and d being reference values), thereby 0 obtaining Idin representing the corrected normalized diameter count-rates.

19. The method as claimed in claim 14 or 15, wherein said correction procedure for said first count-rates employes high energy count-rates from natural sources obtained in the high energy region by means of at least windows which each cover a main energy peak of each respective natural source, and low energy count-rates obtained in the low energy region in a source free detecting device, said count-rates being arranged in a set of equations and processed by means of a regression algorithm calibrated using a Monte Carlo method, whereafter said low energy count-rates processed are compared with said first count-rates resulting in a final correction in said first count-rates.

20. Method for the determination of density, diameter, lithology parameters and elemental concentrations of at least one core sample substantially as described in the description with reference to figs. 1-4 of the appended drawings.

21. Method for the correction of density, diameter, lithology parameter and elemental concentrations of at least one core sample substantially as described in the description with reference to fig. 5 of the appended drawings.