Sand management/control

Nondestructive Core-Strength Tester Uses Steel Ball To Evaluate Hardness

This paper presents results from a nondestructive core-strength-index tester that is less destructive than the Schmidt hammer and less intrusive, easier, faster, and less expensive than the core scratch tester.

This paper presents results from a nondestructive core-strength-index tester that is less destructive than the Schmidt hammer and less intrusive, easier, faster, and less expensive than the core scratch tester. The portable hardness-index tester measures and compares the impact and rebound velocities of a small steel ball after its collision with a rock surface to determine rock hardness, which, in turn, reflects the relative strength of the rock.

Theory and Application of the Index Tester

The index tester (Fig. 1) is a hand-held, electronic, battery operated, spring-loaded device that provides an indirect method to predict rock strength. The Leeb hardness unit is calculated by comparing the impact (Vi) and rebound (Vr) velocities of a small steel ball after its collision with a rock surface (Fig. 2). The impact body rebounds faster from harder rocks than from softer ones, resulting in a higher energy quotient, which, in turn, reflects the relative strength of the rock. Features of the index tester include a liquid-crystal display of the results, internal data storage, and a connection for download of data for future analysis and reporting.

jpt-2013-10-corestrengthfig1.jpg
Fig. 1—Hardness-index tester.
jpt-2013-10-corestrengthfig2.jpg
Fig. 2—Index-tester mode of operation.

Advantages of the index tester include

  • It is portable, for use in the field or core store.
  • It can be set to compensate for angle of impact.
  • Repeatabilty is good.
  • Little damage occurs at the surface of the core because of the low impact energy.

It is, therefore, less destructive than the higher-impact Schmidt hammer and less intrusive, easier, faster, and less expensive to use than the core scratch tester. Fig. 3 compares the minimal impact left on the surface of the core by the index tester compared with the 10-mm channel cut by the scratch tester.

jpt-2013-10-corestrengthfig3.jpg
Fig. 3—Impact of index tester (12 small indent points) on left vs. that of core scratcher (10-mm channel along the length of the slabbed core surface) on right.

The index tester is run at regular intervals along the core surface, with the objective of using the data as an aid in sample selection and then to establish a simple equation for estimating rock strength from the Leeb hardness values calibrated to core-measured rock-strength data.

Practical Application of the Index Tester

Before viewing the core, it is recommended that all available log data and depth-shifted routine-core-analysis data be loaded to a petrophysical database to highlight the best-quality intervals with respect to porosity and permeability. There are a number of published and proprietary log/core strength correlations that can be used to develop a continuous rock-strength model in the absence of core data. Most models involve correlations between unconfined compressive strength (UCS) and logs that are sensitive to rock-strength variations—particularly sonic, density, and porosity. All available generic log-based models should be run over the reservoir intervals and displayed.

Core-rock-mechanics sample selection focuses primarily on the weakest intervals of the core but should include some stronger intervals to improve the dynamic range for log calibration. All available core should be sampled with the index tester at regular intervals, ensuring that the highest-porosity and permeability facies are included. Tests at routine-core-analysis plug points enable integration of results with porosity/permeability data to aid in sample selection. To account for local, small-scale variation in the core, it is recommended to take 12 readings within a limited area of the rock surface at each test site, over an area approximately equivalent to the area of a core plug (see Fig. 3). The highest and lowest readings are excluded, to avoid anomalous data, before taking the average of the remaining 10 readings.

Subsequent to the core viewing, the index-tester results are upscaled to rock strength by use of generic correlations between Leeb number and rock strength and plotted with depth. Suitable sample points can then be selected through integration of all log and core data, core photos, and index-tester results. On receipt of the laboratory results, the index-tester data are calibrated to the field-specific core results and integrated with the core UCS or thick-wall cylinder (TWC) data to provide a continuous estimate of rock strength over the cored interval. Finally, a log-based rock-strength model is calibrated to the core and index-tester data to characterize rock strength in the uncored intervals and wells in the field.

Beware Clay Content

In one sample, there was a very narrow dynamic range on the UCS and index-test results; the core UCS results are in the range of 1,227 to 2,060 psi. The Leeb data, when plotted against the core UCS results, fell below the generic Verwaal and Mulder trend. This interval had significant clay and fines content and exhibited more-plastic, less-elastic deformation on stress loading than cleaner sands. This highlights the necessity for field-specific core calibration of the index-tester data and log-based models. In the untested cored intervals, the index-test data matched the subtle log-based-model trend, thereby increasing confidence in the model.

Beware Grain Size

Index-tester results upscaled using the Verwaal and Mulder correlation can overestimate rock strength in very-coarse-grained sands. The 3-mm-diameter spherical test tip can strike individual large sand grains, resulting in a higher-than-expected Leeb number. The reservoir sands in one sample were weak and coarse-grained, with core UCS values in the range of 521 to 903 psi. The predicted UCS from the Verwaal and Mulder model was significantly higher than the field UCS-test results. Integration of all available log, index-tester, and field rock-mechanics core data resulted in a good match between the core TWC tests and the final log-based model. The calibrated index-tester data support the log-based trend in untested intervals and underscore the necessity for field-specific core calibration.

Caveats

There are several caveats and considerations regarding the use of the index-tester results.

  • The results can be unrepresentative in very-coarse-grained material because the 3-mm-diameter spherical test tip can strike individual large grains.
  • Slab resination can artificially strengthen core if pores become invaded with resin, with the result that the upscaled index-tester data overpredict with respect to the core UCS data.
  • As with core-test data, upscaling from index-tester scale (25 mm) to log scale (1000 mm) can be an issue because standard logs cannot resolve thinly bedded intervals.

Conclusions

Definitive rock-strength data for sand-failure evaluation are available only from tests on core plugs, but coverage is limited and core condition and geometry may preclude plugging. The hardness-index tester is a portable, inexpensive, nondestructive tool that complements but does not replace core-test data. Use of the index tester optimizes core-sample selection and ensures representative results in all cored facies and formations.

The index tester provides an indirect method to predict rock strength. Generic correlations for upscaling the raw Leeb data to UCS are useful as an approximation of relative rock strength before field-specific calibration. Existing correlations were based on limited databases of mixed lithology, with few data points below 1,000 psi. Results from an extended database of cores from sandstone reservoirs worldwide indicate that field-specific calibration is essential, especially in weaker sands where factors such as grain size and clay content have a significant influence on rock strength.

There is a robust relationship between UCS/TWC data and Leeb number over a wide strength range. Calibration with field-specific core refines the relationship so that the index tester can be used to supplement core data points in the calibration of continuous log-based rock-strength models to characterize uncored intervals and wells. These rock-strength models represent key tools in the development of effective sand-management strategies for the reservoir life cycle.

This article, written by Editorial Manager Adam Wilson, contains highlights of paper SPE 158326, “Nondestructive Strength-Index Testing Applications for Sand Failure Evaluation,” by Gillian Daniels, SPE, Colin McPhee, SPE, Philip McCurdy, SPE, and Yelitza Sorrentino, SPE, Senergy, prepared for the 2012 SPE Asia Pacific Oil and Gas Conference and Exhibition, Perth, Australia, 22–24 October. The paper has not been peer reviewed.