LOG PICKING - "BOXING THE LOG"
This article is based on Crain's Petrophysical Pocket Pal" by E. R. (Ross) Crain, P.Eng. Copyright 2018 by the author. Copying or distribution in any form without express written permission is prohibited.

In order to perform a log analysis, it is necessary to read or pick log values in the various zones of interest, and other key locations, such as in shale or water bearing zones. Picking log values and analysis parameters from logs is THE most important step in quantitative log analysis. Mathematics cannot compensate for poor selections. Few comments on this subject are found in service company training manuals. As a result, beginners often find it difficult to start with valid data, or assume the task is easy and requires no thought or knowledge.

Selections should be made on a consistent basis from day to day to assist reproducibility of results. When using digital log data, the digits themselves will be used by the computer program, but the analyst must still pick numerous values by observation of log curves, crossplots, or data listings.

In computer aided log analysis, picks are made continuously with a digitizer or by reading magnetic tapes or discs created when the logs were recorded. Such data tends to be more accurate than hand picked values. Accuracy can be a hindrance on noisy logs, rounded bed boundaries, or in large or rough holes. Some editing or curve shaping may be required prior to digitizing, hand picking data, or using existing digitally recorded data.

To select a log value, it is helpful, especially for the novice, to "box the log". Draw horizontal lines at each bed boundary, at the inflection points on each curve. Draw vertical lines on each curve at the peaks and valleys, thus transforming the log into a series of individual beds with a single specific log reading. For thicker reservoir layers, an average value may be chosen if the log curve does not wiggle too much.

Reading well logs is made easier by "boxing the log". Horizontal lines (bed boundaries) and vertical lines
(log values or log readings) form boxes that bound the data to allow valid data picks. Be sure to honour the "peaks and valleys" rule in thin beds.

With experience, it is possible to simply mark points at the peaks and valleys without drawing horizontal lines, as shown in the lower part of the example. Unless absolutely necessary, values should not be selected on slopes. Slopes indicate transition from one condition, such as porosity or hydrocarbon content, to another. Average values, halfway along the slope may be meaningful, but can also be misleading. Do not select values in thin beds unless you are also prepared to make bed thickness corrections.

Be sure to pick the correct curve, its appropriate scale, and edit any noise or bad hole conditions prior to finalizing values.

Write the log values picked into a table or preprinted form so that you will have a record of the data you are using. Note that very shaly zones are not usually interpreted. Therefore, this data can be left off the table or marked as shale with no data values entered.

Shale Properties:
            Gamma Ray clean line (GR0)         ___22___ api units

Gamma Ray shale line (GR100)    ___105___ api units

SP clean line (SP0)                       ____14___ mv

SP shale line (SP100)                    ____78___ mv

Density shale line (PHIDSH)          __0.03___

Neutron shale line (PHINSH)          __0.26___

Sonic shale line (DTCSH)              ___NA___ us/m

Resistivity shale line (RSH)           ___20____ ohm-m

            Resistivity of  Water Zone (R0)      ___2.7___  ohm-m

Log Data Values for Layers                                                              A: Sh Sd       B: Oil     C: Water

Layer Top                                                                                         __3364_        _3499_    _3512__ feet
Layer Bottom                                                                                    __3406_        _3512_    _3530__ feet
Deep Resistivity                 RESD                                                      ___20__        __100_    __2.7___ ohm-m
Neutron Porosity                PHIN                                                        __0.24__      __0.27_    __0.24_
Density Porosity                 PHID                                                        __0.12__      __0.22_    __0.19_
Sonic Travel Time             DTC                                                         __NA____    __NA__    __NA__ usec/m
Gamma Ray                       GR                                                          __75____    __25__    ___30__  api units
Spontaneous Potential      SP                                                           __55____    __23__    ___15__  mv
Photo Electric Effect           PE                                                          __3.4___     __2.0__    ___2.0__ cu
Caliper                              CAL                                                          __8.5__      __7.7__    ___7.7__ inches

When using computers, log data is usually digitized at an increment much finer than the tool resolution. Thus answers are calculated even on slopes and in thin beds. Understanding such results usually requires some thought. Again, be sure to pick the correct curve from the data file.

General Rules For Picking Log Values

In order to perform a log analysis, it is necessary to read or pick log values in the various zones of interest, and other key locations, such as in shale or water bearing zones. Selections should be made on a consistent basis from day to day to assist reproducibility of results.

In computer aided log analysis, picks are made continuously with a digitizer or by reading magnetic tapes created when the logs were recorded. Such data tends to be more accurate than hand picked values. Accuracy can be a hindrance on noisy logs, rounded bed boundaries, or in large or rough holes. Some editing or curve shaping may be required prior to digitizing, hand picking data, or using tape-recorded data.

To select a log value it is helpful, especially for the novice, to "box the log". Draw horizontal lines at each bed boundary, at the inflection points on each curve. In thinner beds, draw vertical lines on each curve at the peaks and valleys, thus transforming the log into a series of individual beds with a single specific log reading. Pick peaks or valleys in thin beds to get the best possible values.

On thicker beds, draw a line through the average value of the curve. It is necessary to create a new bed or layer each time the porosity changes by 2% (porosity units), or when resistivity or gamma ray change by more than 10% (relative units)

The rule is to draw bed boundaries at the top and bottom of each clean zone, then draw boundaries at the porosity breaks within each clean zone. Finally draw any new boundaries needed to accommodate resistivity changes, which usually represent oil or gas water interfaces. The major interval between the top and bottom boundary of a relatively clean rock is called a ZONE. Each boundary inside a ZONE defines a LAYER of rock. Each layer will be analyzed separately by picking log values for each layer, called "Reading the Log"..

This concept is shown on the log at the right.

For thick layers, pick average values. For porosity, gamma ray, laterolog, and array induction logs, "thick" means more than 6 feet (2 meters). For older style induction logs, "thick means greater than 15 feet (5 meters).

In thick beds, pick average values
 (heavy black vertical lines)  =

Old style induction log, layer roughly
 15 feet (5 meters), pick peaks and valleys;
 other logs, pick averages =

= Thin beds, < 6 feet (2 meters), pick peaks on porosity logs. If there are high porosity streaks, treat them as individual layers. Average values will severely underestimate permeability and productivity. 90% of the production in this zone comes from layer B.

Unless absolutely necessary, values should not be selected on slopes. Slopes indicate transition from one condition, such as porosity or hydrocarbon content, to another. Average values, halfway along the slope may be meaningful, but can also be misleading. Do not select values in thin beds unless you are also prepared to make bed thickness corrections.

PICKING FLUID, MATRIX, and SHALE PARAMETERS
The methods for  selecting petrophysical parameters depends on whether knowledge of fluid, matrix, or shale values is needed.

Fluid values for various analysis methods are generally obtained in a laboratory environment and adjusted for temperature, pressure, and salinity as required. They cannot generally be picked directly from logs.

Matrix rock values are normally available from handbooks or data tables. The numbers usually represent log readings for pure minerals, which rarely exist in real situations. The values may also be found by inspecting logs if relatively pure, zero porosity zones are present. Some crossplots may assist in finding matrix parameters.

Due to varying shale compositions, shale values are not as well known or as constant as for other rock minerals. They are often found by inspecting logs in a shale bed near the zone being interpreted. Some crossplots may assist in finding matrix parameters.

In order to pick a parameter, the expected values must be known approximately. Only then is it possible to determine if the value seen on the log or the crossplot is reasonable and representative of the parameter required. This may involve evaluating several wells to gain confidence in making assumptions.

Suggested methods for selecting parameters through log inspection are illustrated below and in the following discussion.

1. Shale resistivity is the average value of the deepest resistivity curve reading in shale, 20 feet or more thick, below the zone in question. If the well does not penetrate a shale below the zone of interest, or if a major geological boundary is crossed, use the shale immediately above the zone.

2. Resistivity in a water zone is the lowest value of the deepest resistivity curve reading in a water zone, 20 feet or more thick, below the zone to be interpreted. The value may be slightly to 2 or 3 times too high. It is used to determine water resistivity for water saturation calculations.

3. It is usually impossible to obtain matrix values from observation of log data alone. Matrix values for the sonic, density, and neutron logs are used to correct for the effects of the varying lithology. Find the lowest consistent value of sonic travel time, lowest density, porosity, (or highest density), and lowest neutron porosity in the zone to be interpreted. If these values are close to the expected matrix value for the known lithology, they may be used with caution. If lithology is unknown, start with pure mineral values from tables.

4. Shale values for sonic, density, and neutron are determined from the average value of logs in shales, 20 or more feet thick, below the zone to be interpreted. This applies to clean logs without skips, spikes, and rough or large boreholes. Caution should be used since shale properties can vary widely within a short interval. Data is used for shale corrections to porosity calculations. Therefore, corrections may be inaccurate if shale properties vary or are poorly chosen.

5.Gamma ray readings in shale are usually reasonably high (100 - 200 API units and lower in sandstones and carbonates (10 - 50 API units). SP scales vary widely, but deflections to the left usually represent sandstones and deflections to the right mean shale, except in shallow fresh water sands. The SP is usually featureless in carbonates.

6. The gamma ray and SP clean sand and shale lines are required to allow us to find the shale volume for use in shale corrections to porosity calculations. To determine the clean line value, find the cleanest or least shaly zone in the entire well. Lower this value to suit the known shale content in other zones. Caution - never push the clean line into more than 5% of the data points. To find the shale line, draw a line through the average data value in thick shale zones. Do not include very radioactive zones which are generally caused by uranium, and not shale minerals. Up to 10% of the data points may be above the shale line.

EXAMPLES OF READING LOGS

 
Gamma ray and SP clean lines (GR0 and SP0) and shale lines (GR100 and SP100) picked for several zones. Resistivity of shale (RSH) and resistivity of water zones (R0) are also shown. Note variations with depth due to changing geological situations.

Sonic travel time in shale (DELTSH) also varies with depth.

Density, neutron, and gamma ray shale properties vary with depth
As do the neutron porosity of shale (PHINSH) and density porosity of shale (PHIDSH).

If base line methods are difficult, certain crossplots may be helpful. Some analysts prefer the crossplot method although it requires an extra computer step and is not appropriate for visual or quick look interpretation. Crossplots also mask variations related to depth, so appropriate zoning is required.