Considerations and practical applications
By Marvin F. Glotfelty, RG
An envelope of filter pack sand (Figure 1) is necessary for most wells in unconsolidated or friable formations to prevent native sand from flowing in through well screen slots during pumping.
Invasion of fine native sand will damage pump equipment and clog the water distribution system. To provide for the inflow of groundwater while preventing fine sand from migrating into the well, the filter pack should achieve several objectives:
- Prevent native sand from migrating into the well during
- Allow energy to be conveyed through the sand envelope during well development
- Stabilize the well structure.
The third objective is easily achieved with an adequately hard filter pack material that has been properly placed to fill the annulus around the well screen. The first two objectives can also be achieved by selecting a filter pack with these characteristics:
- Well sorted (poorly graded)
- Well rounded and spherical
- Made of silica-rich minerals (such as quartz or feldspar) or other materials that are hard and inert
- Proper grain size in relation to the adjacent formation.
Water wells are designed for a broad variety of uses, so no single material source is appropriate for all wells. Most filter pack material is quarried sand, but manufactured glass beads are also available from several sources (Figure 2E).
The desired filter pack characteristics listed above are critical for some wells, but somewhat trivial in other cases. For example, the filter pack shown in Figure 2A would be too angular for many agricultural production wells but would be acceptable for monitoring wells from which large pumping rates are not needed.
On the other hand, the filter pack shown in Figure 2C would be acceptable for most agricultural wells, but would not be a good choice for some monitoring wells because the dark minerals in it could impact the well’s water quality if metals like iron or magnesium were solubilized.
Sieve Analysis and Filter Pack Design Calculations
There are numerous references for calculating the optimum grain size for a water well’s filter pack. Table 1 isn’t comprehensive but lists several of the more recent and well-known sources. Formulas from different sources vary but are generally quite similar.
Filter pack design generally starts with a sieve analysis of drilled cuttings from the borehole to provide a site-specific basis for the design. The process of conducting a sieve analysis involves passing a weighed sediment sample through a series of stacked sieve screens with various mesh sizes (Figure 3).
The coarsest sieve screen is positioned at the top, with finer-meshed screens farther down, so the sample will be segregated by grain size as it is retained on the different-sized screens. The proportional amounts of the sample that are retained on each sieve screen are then weighed to determine the percent of each grain size.
The sieve analysis results are typically plotted as percent by weight on the vertical axis versus grain size on the horizontal axis. Sieve analysis plots can be presented as percent retained versus grain size (Figure 4, top) or as percent passing versus grain size (Figure 4, bottom).
The two types of sieve analysis plots provide the same information, but their vertical scales are reversed because a sample that is 60% retained by a particular screen also has 40% passing that screen. The common nomenclature for grain size diameter is Dxx. Thus, 70% retained would be referenced as D70 (Figure 4, top), which is equivalent to 30% passing that would be designated D30 (Figure 4, bottom).
Karl Terzaghi’s 1943 publication, Theoretical Soil Mechanics, introduced the principle that filter pack sand should be four times coarser than the D85 retained formation sediment (Table 1).
Terzaghi’s formula was actually developed for drain systems to prevent internal erosion called “piping” within earthen dams. Impounded water flowing through a drain beneath an earthen dam is very different than groundwater flowing into a pumping well. However, during the past 77 years, multiple well designers have applied variations of Terzaghi’s formula
to successfully design filter packs for thousands of wells.
A confusing aspect of filter pack design is the fact that we commonly refer to well screen slot sizes in terms of inches, but sediment grain size in terms of U.S. Standard mesh size. Table 2 provides a correlation between U.S. Standard mesh numbers and slot sizes.
Sphericity of Filter Pack Grains
A desirable characteristic of filter pack is for it to be well rounded, so that flat surfaces of the sand grains will not lay against one another and disrupt the filter media porosity. The filter pack should also be spherical rather than elongated in shape.
The sphericity and roundness of sand grains were categorized by Krumbein and Sloss in their 1963 publication, Stratigraphy and Sedimentation. Filter pack should have a Krumbein and Sloss scale value (Figure 5) greater than 0.6 for both roundness and sphericity.
Even well-rounded sand can sometimes have mediocre sphericity characteristics (Figure 2B and 2C). I encountered this many years ago when a filter pack sand I had specified was expected to be more than 90% retained by the well screen. However, when a handful of sand was poured on the face of the well screen prior to its installation, more than 60% of the sand passed through.
The sand was rejected, and we sent a sample of it to a geotechnical lab for re-analysis. The supplier and the driller also analyzed duplicate samples, and all three sieve analysis results indicated that only 10% should have passed through that screen slot size. We then realized that the equally spaced slots in a sieve screen are not entirely representative of the elongated slots in a water well screen. We adjusted the grain size to accommodate this discrepancy, but learned the lesson that differently shaped screen slots provide variable results when sand grains are less spherical.
Well Design Lessons Learned
About 30 years ago, I undertook the process of designing my first public supply water well, and thus embarked on a lifelong well design learning curve that continues to this day. For the first few years of my career, my filter pack designs were based strictly on published sieve analysis formulas (Table 1). The grain sizes for each filter pack I specified were generally 6 to 8 times coarser than the D70 grain size of the formation. I learned many fundamental lessons during those early years, and fortunately, none of the wells I designed had inadequate groundwater production or sanding problems when they were installed.
A few years later (about 25 years ago), I was involved with a well installation project at which very fine-grained “sugar sand” had been encountered. The formation was unstable flowing sand, which unfortunately caused the driller to lose a string of collars in the borehole, so the rig was moved over to re-drill the well.
The screen type preferred by the well owner (louvered screen) had a slot size limit of about 0.050-inches, so the filter pack had to be 8-12 mesh or coarser. Thus, I was concerned about sand invasion problems in the completed well. Even though the filter pack grain size was much coarser than would have been determined by the sieve analysis formula I’d always used, the well produced a respectable 1200 gpm with no sand production. This experience made me wonder about the applicability of the formula upon which I had relied for so many years.
Art of Filter Pack Design
The filter pack design formula (in its various forms) has been used by numerous well designers in many different geologic environments for almost eight decades, so it is reasonable to consider it as a tried-and-true approach for filter pack design.
On the other hand, every single element of well design (including filter pack selection) involves skillfully embracing the data that we can collect and consider, while also recognizing the immeasurable areas where we have no information. Water well design involves an interaction with Mother Nature, so we should always remain mindful that our understanding of the hydrogeologic environment is actually quite limited.
For example, we may determine the D70 grain size of our cuttings based on sieve analysis, but we shouldn’t forget that our cuttings sample represents about 10 vertical feet of borehole depth. In that 10 vertical feet, a stratified formation could vary in numerous different ways. Thus, the D70 grain size may not be as representative of the site-specific conditions as we may have thought.
Over time, seemingly minor or unforeseen aspects of the filter pack design can significantly influence the well performance. During a well’s operational life, portions of the filter pack may become clogged by scale accumulation (Figure 6), which can occur in very localized intervals of the well screen.
Even without clogging from scale, some wells have a tendency for focused sand invasion at discrete locations, due to high entrance velocities at those depths. Figure 7 shows video images from a well as it is being pumped, with the upward flow of sand observed at 823 feet. Sand invasion does not occur throughout the entire screened interval, but only locally at a few locations with no sand entering below 830 feet.
There is no reason to question the validity of the tried-and-true sieve analysis process for filter pack design (Table 1), and it should be continued. However, it is an analytical technique that should—as with all the other aspects of well design—be used in conjunction with an appreciation of all the hydrogeologic influences (known or unknown) that will impact our well’s performance.
Filter pack design is an important component of well design, but as with every other aspect of well design, it should be conducted with an awareness of our data limitations. We should maximize the information available to us, while respecting the significance of those things we cannot measure or quantify.
With all our state-of-the-science analytical techniques, well design remains an art, which is augmented by the well designer’s scientific judgment, experience, and common sense.
Marvin F. Glotfelty, RG, is the principal hydrogeologist for Clear Creek Associates, a Geo-Logic Associates Co. He is a licensed well driller and registered professional geologist in Arizona, where he has practiced water resources consulting for more than 35 years. He is author of The Art of Water Wells (NGWA Press, 2019) and was The Groundwater Foundation’s 2012 McEllhiney Lecturer. Glotfelty can be reached at firstname.lastname@example.org.