For nearly six decades, hydraulic fracturing (fracking) has been used to reach deep petroleum deposits. However, it is only of late that this method has been exploited to reach natural gas reserves, which are typically located closer to the ground surface. Fracking for natural gas deposits requires a small degree of vertical drilling, followed by horizontal drilling toward the gas reserve. Proppants (finely sized particulates), chemicals, and water are injected at high pressures to blast open shale rock and allow the trapped gas to escape. The proppants (propping agents) “prop” the fracture open to allow a path for gas flow. Proppants can be sand (sandstone or quartz/silica), ceramic crystals or resin-coated sand. Since the latest large-scale discovery of several rich natural gas reserves, usage of fracking has grown to such a rate that in 2013 20 million tons of frac sand was used in North America alone. Several sandstone deposits containing sand ideal for fracking are located in Wisconsin, Minnesota, Missouri, Illinois, and Texas.
Proppants have to satisfy a number of very stringent specifications for, among other things; size distribution, and two particle shape parameters. Smooth rounded shapes and uniform sizes maximize the flowability, or conductivity, of the gas from the reserve to the surface. The American Petroleum Institute (API) dictates these specifications, current API RP 19C (ISO 13503-2:2006, Identical).
The present particle shape standards stay the same as those established in the 1960s by Krumbein and Sloss, who developed a 2D silhouette chart of differently-shaped sand particles, based on their terms for Roundness in the X axis and Sphericity in the Y axis.
Krumbein-Sloss Shape Factors
According to API RP 19C, the Krumbein-Sloss shape factors for sand particles are determined by a technician observing a minimum of 20 grains of a representative sample on a manual optical microscope, and subjectively allocating a Sphericity and Roundness value to each by visually comparing the particles to the images in the table and allocating values listed on the chart.
Calculations of these shape values, which were based on rather outdated formulae, are never conducted. However, this present version of the API standard states that techniques using digital technology are currently available and appropriate, meaning that dynamic image analyzers like Microtrac’s PartAn3D are suitable for establishing these values. Preferred procedures are not mentioned for using these image analyzers, nor are the parameters reported. The Krumbein-Sloss Sphericity and Roundness values are actually measures of form and surface roughness respectively. This means that the PartAn3D parameters reported for Sphericity, Circularity and Ellipse Aspect Ratio could be used for form, and Solidity, Convexity, and Concavity could be used for surface roughness values.
These shape parameters are distributed across a scale from 0 to 1.0. A value of 1.0 for Sphericity, Circularity and Ellipse Aspect Ratio would indicate that the particle is a perfect sphere. Values of 1.0 for Solidity and Convexity would point to particles with a completely smooth surface without any indentations, and a value of 0 for Concavity would point to a particle with a fully smooth, convex surface.
The Microtrac PartAn3D particle image analyzer was used to measure a sample of processed fracking sand. The PartAn3D is a dry dynamic particle image analyzer. The sample is deposited in a hopper which feeds into to a vibratory feeder, which supplies the particles to a point where they drop, due to gravity, through the sensing zone of the analyzer. A high-speed strobe light is used to strike the particles from one side of the sensing zone, while a digital camera is used to capture the back-lit images of the particles. The images are stored in a video file for analysis. Thirty-two different shape, size, and intensity parameters (13 size, 17 shape, 2 intensity) are measured for each particle in the image file and reported as tabular and graphical Number or Volume distributions, together with user-selectable summary data like percentiles, means, and several more. The image file is also viewable and can be sorted in ascending or descending order on any parameter. The saved video file can be re-run later under various measurement conditions.
Both size distributions and the various form and roughness shape parameter distributions were measured and reported, in tabular form and graphically, together with summary statistical data. The Microtrac PartAn3D particle image analyzer has several unique benefits, compared to sieve and microscopic analysis, for measuring the shape and size factors required by the API standard for proppants. They are listed in the table below.
|Ease and accuracy of measurement
||Thirty different size and shape parameters reported in one automated measurement of up to hundreds of thousands of measurements; each individual particle, taking only a few minutes to complete.
||Fifteen-minute manual ensemble sieve analysis, subject to operator error and worn sieves, giving one size, equivalent spherical diameter. Up to a day-long subjective manual microscopic analysis to get representative sample, giving two estimated mean shapes.
||Fifteen different size, twelve different shape and two translucence parameters reported as graphical and tabular number or volume distributions, with means and much more summary data, plus a two-parameter comparative scatter diagram isolating the relationship between any two parameters and a stored re-measurable sortable viewable printable image file of each particle.
||One size analysis as a mass distribution with mean values and a few other summary data. No data is given for each individual particle, only for the entire ensemble sample. A subjective operator-dependent estimate of two shape factors, as mean values of the entire population measured.
||PartAn3D is the only image analyzer with a patented 3-D mode of analysis. All three major axes of each particle are measured and reported. The volume of each particle is calculated as the product of its length, width and thickness. This provides a far more accurate volume size distribution than equivalent spherical diameter calculated as an ensemble volume distribution. Shape parameters are ratios of various size parameters, which gives the PartAn3D the most accurate and automatic shape data quickly and easily measured.
||Sieve analysis gives only an equivalent spherical diameter size parameter. Manual microscopic analysis gives only operator-estimated means of a 2-D analysis for two shapes.
Figure 1. Partial rows of images from the video file of a processed frac sand. In the PartAn3D 3-D mode, each row represents a series of photos of one particle in random orientations as it tumbles through the sensing zone. The top row shows a large particle made up of agglomerates of individual grains. In the bottom row, one image is in an orientation which shows the particle to be an agglomerate of 5 grains. Agglomerates are likely de-agglomerated under the high pressure and abrasion which occurs during the fracturing process, as they are similarly during the physical pounding during a sieve analysis. This indicates that image analysis, run on a fresh sample, would report the degree and type of agglomeration in the process product, and then, after a short period of vigorous shaking of the sample, which is always recovered after an analysis, the de-agglomerated sample could be measured, representing the sizes and shapes that would exist when packed in the fracture.
Figure 2. This is the Scatter Diagram display illustrating Sphericity on a scale of 0 to 1.0, where 1.0 represents a perfect sphere, plotted on the x-axis on top, and Convexity, also on scale of 0 to 1.0, where 1.0 represents a perfectly smooth particle with no angularity or surface roughness, plotted on the y-axis to the right. Any parameter can be displayed, along with its summary data to the right of the y-axis. The PartAn3D Sphericity and Convexity correlate with the Krumbein-Sloss Sphericity and Roundness values respectively. Recall the API standard calls for these values to be 0.6 or greater. The PartAn3D mean values here are 0.93 and 0.99. The darker the blue in the Scatter Diagram, the higher the concentration of particles.
Figure 3. This is the X-Y Graph plot of four different size parameters: length, width, thickness and area-equivalent diameter for the processed frac sand sample, with size channel edges set at the API standard required sizes, plotted as volume distributions with tabular data listed to the right. Any of these parameters can be correlated with the standard sieve distribution, and once correlated, reported as the % values of a sieve measurement for each size fraction. User-selected summary data are shown in the lower right window.
Figure 4. This is the upper right corner of the View Particles display, showing all 2-D and 3-D parametric data for the particle chosen by the user, highlighted inside the blue box in the frame window.
This information has been sourced, reviewed and adapted from materials provided by Microtrac.
For more information on this source, please visit Microtrac.