Return To Handbook IndexRoss Crain's Resume
W
ELCOME  TO
CRAIN'S PETROPHYSICAL HANDBOOK

Please be fair to the author. Pay your Shareware Fee HERE and receive a copy of CPH by download.

CHECKING LOGGING PROBLEMS

       Depth Shift Problems        Skips, Noise, Spikes        Rough, Large, and Salty Boreholes

Depth Shifts
Logs run on separate passes can easily be off-depth from each other. Choose one curve, usually the primary porosity curve (eg. density or sonic or gamma ray), and make it the reference curve. Compare all bed boundaries on each curve to the reference curve, and build a table of depth adjustments over the interval to be analyzed. Most computerized log analysis packages can stretch or squeeze log curves based on such a table, or allow interactive picking of the table from screen images.

Curves that are run on the same pass can also be off depth with each other. Some curves are memorized by the recording equipment because they are actually recorded a few feet above the lowest curve on the logging tool. The operator can set the memory distance incorrectly or the electronic memorizer can malfunction. Even computer controlled logging does not eliminate this problem.

Should the logging tool pull tight in the hole, it may slow down or even stop, while the cable keeps moving and stretching. Since the cable movement drives the recording camera, curves will be recorded off depth during this period. As a result of the memorizer, each curve will be off depth at a different location on the log.

After the tool pulls free it will move up the hole quickly while the logging cable driving the recorder will not. Therefore, the logs are again off depth, but in the opposite direction. This problem can only be adequately resolved if a curve exists which did not pull tight. Other curves can be stretched and squeezed to match the reference curve by computer programs or by approximate depth shift tables constructed by hand.


Skips, Noise, and Spikes
Cycle skips and noise are normally related to sonic logs, but can occur on any log curve. To edit, draw a smooth log curve ignoring the spikes, following an imaginary base log located beneath the noise, as shown below.


Editing sonic log skips

At times, a smooth log curve may not exist. Therefore, it is necessary to review an offset log or another curve from the same well. Note that, the edit gives a sonic porosity of 5% - 10% instead of 25%. It indicates a major difference and is more probable for this particular zone.


Editing density log for bad hole condition

CAUTION: Some noise may be the result of thinly bedded porous layers, coal, or rough hole effects. Coal spikes should be identified as such, and bad hole effects discounted. Some bad hole is caused by breakout of the wellbore at natural fractures. The density log is the most strongly affected curve. The density log porosity should not be used as an indicator of reservoir volume, but the location of the fractures should be noted.

If the sonic log is edited, the integrated transit time curve must also be corrected.


Rough, Large, and Salty Hole Effects
All logs are affected to some degree by borehole size and environment. Application of environmental corrections by computer programs may reduce the need for editing. If no computer is handy, you may have to do something yourself. Several examples are cited here. The log below shows a resistivity log which is severely affected by very conductive mud in the borehole. The dual laterolog would have been more suitable in this instance. Since only the dual induction was run, its readings must be used.


Dual Induction Log in salt mud (left), good log from offset well (right)

As a result two problems occur. Conductive mud causes the induction log to read too low due to invasion and borehole effect. This can be corrected by using special charts for invasion and borehole corrections. These charts are unique to particular tool types and can be obtained from service companies. Borehole correction charts are available from each service company and may be incorporated in computer aided log analysis programs. Since each tool from each company requires its own individual correction chart, the charts in the computer program may contain only a generic correction.

The second problem is the hashy nature of the log in low porosity zones, caused by variations in borehole effect. An edit is required to select a reasonable resistivity value and is shown on the log. The offset log, as well as the gamma ray, sonic and density neutron curves are used to provide the proper amplitude and shape. This is obviously a very serious change of log values and all calculated results should be used with extreme caution.

CRAIN’S OPINIONATED OPINION #1

Borehole size, mud weight, and mud salinity corrections are usually useful corrections to make.

Invasion correction charts are almost universally useless. The borehole and resistivity regime needed to satisfy the available charts are almost never met in real rocks. In particular, the published correction charts are designed for mud filtrate invasion into water zones and NOT for hydrocarbon bearing zones. Induction log invasion corrections in a fresh mud system in a hydrocarbon zone will reduce resistivity instead of raising it. If a computer program has both environmental and invasion corrections, leave the environment corrections turned ON but the invasion corrections turned OFF!

Borehole corrections for gamma ray logs are useful. Corrections for most neutron logs over-do the work and raise porosity too much. Leave neutron corrections turned OFF.



Modern sonic logs are normally unaffected by hole size. At times, the hole may be too large and the tool cannot find a refraction path that will deliver sound to the receivers. The log may be edited by comparison to offset wells, or reconstructed from calculations based on other logs. Older style, single transmitter, sonic logs have a spike at each change in hole size which must be edited manually.

All pad type devices, such as density, sidewall neutron, microlog, proximity, and dipmeter logs are ineffective in large holes, where the hole diameter is beyond the reach of the pad. Tools will either read the mud value or jump from high to low values due to intermittent contact with the borehole wall. Log data from these curves must not be used in this environment.

In holes ranging from bit size to approximately 30% of the pad extension, the tool response is probably correct. From 30% to 90% extension, logs may appear reasonable, but are influenced by mud between the pad and borehole wall. A similar effect may arise in thick mud cake where the hole size is smaller than the bit size.

Rough or rugose holes will also leave excessive mud between the pad and borehole wall. This effect may not seem too noticeable on most logs, since the caliper may appear smooth. On the density log, the correction curve will be abnormally high in such zones. This problem cannot be detected on most other logs. Therefore, caution is recommended. The MSFL does have an apparent mud cake thickness curve (Hmc) which will be abnormally high in such conditions.


 Density Logs in rough hole

We know these logs are inaccurate because the apparent porosity is too high compared to other sources of information such as the sonic and offset wells. As well, the density neutron crossover infers gas, yet this is not indicated by other logs or well data. The porosity is abnormally high for a Devonian carbonate and the density correction curve suggests caution. For reservoir evaluation, the density data should be ignored. For seismic uses, an approximate density would be estimated based on lithologic description and offset well data.
 

Copyright E. R. (Ross) Crain, P.Eng.  email
Read the Fine Print