“MAGIC MOVEMENT” – UNDERSTANDING HORIZONTAL DISPERSAL OF

TREATED EFFLUENT BENEATH SAND-BASED, ONSITE, WASTEWATER

TREATMENT SYSTEMS

Donald M. Prince*

ABSTRACT
Innovation and third-party testing & certification have given the onsite wastewater industry a
number of proprietary, passive, sand-based, advanced treatment system technologies which
process wastewater in markedly different ways than conventional or mechanical-based,
secondary treatment systems. One characteristic is the ability of these Combined Treatment &
Dispersal (CTD) systems to transmit the treated effluent horizontally, significant distances
from the point of release in the specified sand and soils beneath the system. If embraced, this
innate feature allows CTD system designs and configurations which can bring significant
savings on equipment, construction and ongoing maintenance costs. Water table mounding in
restrictive soils is minimized when treated effluent is distributed laterally in the specified sand
media and transferred to the underlying soils predominantly via capillary forces. Using a
demonstration model and citing previous work, this paper aims to show that wastewater
systems with intermittent flows, using certain, specified media and performing a consistent,
certain level of biological treatment, will achieve consistent, horizontal movement of treated
effluent to the extent needed for absorption into the native soil.

INTRODUCTION
The forces which govern horizontal movement of treated wastewater are complex and varied.
Additionally, the treatment performance of the system must be at a level and consistency to
prevent biomat clogging of the specified sand and soils used for dispersal beneath the
proprietary units of these systems. Typically, sand-based CTD systems perform treatment to
better-than secondary treatment levels. Gravity, capillary forces (3-dimensional force pulling
in all directions), the boundary and transition from the specified media (medium-to-coarse sand
with limited fines) to the typically finer native soils, all contribute to the quantifiable
phenomenon of passive lateral movement of treated effluent in these systems.
Capillary Forces: Most people are familiar with the forces of capillary draw where the most
common illustration is the varying levels water is drawn vertically up narrow, transparent
glass tubes from a reservoir. The level of water (capillary rise) differs depending on the
diameter of the tube, with the narrowest tube having the most draw (negative pressure) and
highest level of water. The tubes roughly represent the capillary forces within the soils,
where void spaces within the different materials create channels. Coarse materials have large
channels which create a small capillary force and very fine materials have very small
channels which create a large capillary force.

* Donald M. Prince, Technical Advisor,

Presby Environmental, Inc., 143 Airport Road, Whitefield, NH 03598
PH (603) 837-3826; FAX (603) 837-9864; Email: don.prince@presbyeco.com
1
Figure 1: Rise of water in capillary tubes of various sizes at hydrostatic equilibrium. Source:
Lu, N., Likos, W. (2004)

Figure 2: Maximum Height of capillary rise in capillary tube or idealized soil comprised of
uniform spherical particles. Particle size is delineated in terms of soil type for comparison.
Source: Lu, N., Likos, W. (2004)

The horizontal component of the capillary force has the ability to pull liquids through sand or
soil at a much greater distance than the capillary rise, because it is not pulling against gravity.
Sherman, K. (2018) states, “This capillary attraction for water is exerted in all
directions…Capillary water may also move horizontally in soil. This movement is not
influenced by gravity, however, because the force of gravity does not have a horizontal

2
component.” An indication this horizontal movement of capillary water should be significantly
greater than the vertical rise is a statement by Dan Allen: “Apparently, it is not unusual for the
horizontal hydraulic conductivity to be up to as much as ten times the vertical hydraulic
conductivity.” Allen, D. (1980)

Soil Material Capillary Rise hc

(in.)[mm]

Coarse Sand 4.92 [125]

Medium Sand 9.84 [250]

Fine Sand 15.75 [400]

Silt 39.37 [1000]

Table 1: Established data for capillary rise in various soils.

Source Heath (1982), Sherman (2018)
Capillary Fringe: It has been shown through dye tests that when water is introduced from
above to an unsaturated sand media with an induced water table and capillary fringe at the
bottom, the water moves mostly vertical through the unsaturated media. It then radically
changes direction to move laterally when encountering the capillary fringe without intersecting
the water table at least temporarily. A demonstration of this phenomenon was documented in
a paper: Lateral Movement of Water in The Capillary Fringe under Drainfields, Amoozegar
(2006). The second of three tests described in the paper is remarkable in that it tested two
plumes simultaneously. The two plumes moved vertically with some horizontal spread in the
unsaturated sand media and then moved horizontally along the capillary fringe. The two plumes
did not intermingle, but instead remained distinctly separate as they moved along the capillary
fringe, one below the other (see Figure 2). In the conclusion the paper states, “In all cases, the
dye solutions…moved laterally within the capillary fringe without intersecting the water table.”
The conclusion also noted that the capillary fringe may remain aerobic.
A capillary fringe is associated with a water table which should not exist at the specified
sand/native-soil interface. However, it is not unreasonable to anticipate the possibility of
capillary fringe-like characteristics associated with even temporarily saturated conditions in
this area (see Discussion for pertinent observation of sand box testing and saturated or nearly
saturated flow).

3
Figure 3 – Drawing and photograph of Test #2 showing two plumes moving laterally through the
capillary
fringe without intersecting the water table. Amoozegar (2006)

When treated effluent leaves the proprietary component of a sand-based treatment system into
the specified sand, it moves mostly vertically. There is a relatively small capillary suction in
the medium-to-coarse specified sand which will pull it horizontally from the point of release,
however, with only minor influence on lateral movement. The treated effluent moves vertically
to the sand/soil interface. In soils where lateral movement of the treated wastewater would be
necessary to meet dispersal area requirements, the native soil is typically finer and less-
permeable than the specified sand of the treatment system. At the sand/soil interface beneath
the point of release, a portion of the treated wastewater is transferred to the smaller pores of
the native soils predominantly through the forces of capillary draw at a rate typically less than
being delivered to the to the treatment system. The balance of the treated effluent is transported
laterally via forces similar to the capillary fringe above a water table. However, these forces
are dependent on the intermittent flows introduced by the treatment system and other transient
flows such as precipitation. The capillary fringe forces at the sand/soil interface exist due to a
combination of factors, including the horizontal capillary draw in the soils directly below the
System Sand and the difference in permeability between the System Sand and those same
underlying soils, causing a slight gradient in flow sloping away from the area below the point

4
of release and associated capillary fringe-like characteristics above the flow gradient at the
sand soil interface.

Abrupt Soil Textural Changes: When the texture of soil changes abruptly, it can influence
the way and direction water will move through them. Dr. Walter Gardner states, “In a layered
soil, water will not move into a different textured soil until saturation takes place and gravity
affects water movement.”- Gardner (1988) Dr. Gardner made a series of films illustrating the
radical changes in flow characteristics when layers of different textured soils are encountered.
Concerning sand-based, CTD systems, the sand/soil interface in the vast majority of
installations constitutes just such an abrupt change in texture to create lateral movement.
Moreover, the soils that would require the most horizontal movement, clay soils requiring a
large dispersal area, have the most textural difference from the specified sand used in these
systems. In one of Dr. Gardners films, he shows how sand channels connected to the surface
influence water movement into and through these inclusions (see illustration of three
screenshots from the film). When specified sand CTD systems are vented, they are effectively
connected to the surface and benefit from this preferential flow through the sand media.

Figure 4 – Stills of a film showing preferential flow to a sand inclusion connected to the surface
and the opposite for a sand inclusion not connected to the surface.
Illustration -Hsieh; Gardner
(1959)

Regulators have voiced concerns that allowing CTD system designs which utilize significant
lateral dispersal beyond the area of treatment and effluent release, will eventually compromise
the ability of these systems to treat wastewater to secondary treatment standards due to the
development of saturated conditions in an area of the system that is presumably unsaturated
and designated for aerobic treatment: the 6 inches of specified sand beneath the proprietary
units. For this reason, we chose to load the sand box model aggressively to simulate a worst-
case scenario, but not so aggressively as to abandon any references to secondary treatment.
The amount to load the sand box model was determined based upon actual proprietary testing
at the Massachusetts Septic System Testing Center (MASSTC) in Barnstable, MA. The Presby
Simple-Septic Wastewater Treatment System was tested under the NSF/ANSI Standard 40,
Class I certification for secondary treatment systems in 2013/14, including winter months. This
test was unique in that it tested the proprietary pipe system in serial distribution, whereupon
the 450 GPD actual effluent flow to the system was fed to the first, 30-foot row of a 5-row bed
throughout the test. This system met secondary treatment levels from the beginning of the test,
which is important to note because for many days, weeks or possibly even months, the system
used a single 30-foot row to facilitate treatment. This system contained a total of 150 feet of
proprietary pipe which was loaded at 3 Gallons Per Day (GPD)/FT for the whole system.

5
 During the period it was treating and dispersing effluent in a single row, it was loaded at 450
GPD in 30 feet or 15 GPD/FT. For our sand box test model, the pipe length is approximately
1/3 foot, so these tests will load at an equivalent 5 GPD.

Figure 5 – Testing configuration for NSF/ANSI Standard 40 certification in serial distribution

Figure 6 – Raised Connections are used in the MASSTC test system to facilitate serial
METHODS AND MATERIALS
As a means of demonstrating lateral movement of secondary-treated wastewater, released from
a proprietary pipe in a specified sand media on native soil, a sand box was created using
transparent polycarbonate panels to show a simulated cross-section. The box’s dimensions are
described in relation to the pipe which would have its length along the contour, but in this test
apparatus would be the shortest dimension: approximate dimensions are 4.5 inches in length;
80 inches wide; and 24 inches thick (height). PVC well-screen cut in half lengthwise was
placed at the bottom along the length of the box for drainage to two holes, one at each end.
Filter fabric was placed over the well-screen to prevent sand from migrating between the well-
screen and the bottom of the box. After the first test was made, modifications to the sand box
were made due to leakage and structural issues: Reinforcement struts were added (two in front,
one in back) to counter deflection of the polycarbonate panels; waterproofing tape was used on
interior seams and molding installed on the front only to reinforce the edge of the polycarbonate
panel. Also, two, 2-inch diameter, slotted standpipes were added, one on each side of the test
model and covered with filter fabric, to facilitate drainage if saturated water travelled to the
side on a layer of soil above the bottom of the box.

Figure 7 – Sand box for each test before addition of sand and soil
For both the fine sand and the silt tests, the effluent loading and dosing was the same, five (5)
Gallons Per Day (GPD) and three (3) doses per day of 1-2/3 gallons each. Given the variable
nature of septic system use, a uniform or rigid dosing schedule is not required for these tests.

7
A tracer dye was used on test #1 but it was ineffectual at showing the characteristics we are
interested. I probably used the wrong type of dye. No dye was used on test #2.
Test #1 involved the testing of fine, bank-run sand (less permeable than the specified System
Sand placed above the fine sand) was uniformly packed at the bottom of the box to a height of
8 inches over the top of the well-screen (approximately 2”) to simulate native soil. The
specified System Sand was uniformly packed to a height of 6 inches, whereupon a proprietary
(Advanced Enviro-Septic) pipe section, (approximately 4 inches long, outside diameter of 12
inches and modified by covering the ends with thin polycarbonate sheets with the top section
of pipe removed to facilitate loading clean water). This model simulates approximately 1/3 of
one-foot of a treatment and dispersal system bed length.

Figure 8 – Dimensions of test materials in sand box test model (Test #1)

Test #2 involved testing a silt soil material (more restrictive than fine sand) underneath and
over the specified System Sand and incorporating level, system sand extensions on both sides
of the proprietary pipe and larger system Sand treatment area. There were modifications to the
sand box to make it leakproof and strengthen it. See comments above for a full description.
There is 7” of silt below the specified sand. There is a minimum of 6” of specified System
Sand in the extensions and a minimum of 6” of specified System Sand below and on either side
of the proprietary pipe. A minimum of 6” of silt was placed over the System Sand extensions.
The top of the pipe was removed to facilitate dosing clean water to the model.

8
Figure 9 – Illustration dimensioning materials for testing (Test #2)
Sieve Analyses:
The specified System Sand is the sand component in contact with the proprietary system and
is typically medium to course sand with limited fines (material passing the #200 sieve). Same
material specified for NSF/ANSI certification. An example specification is concrete sand
which meets the ASTM C-33 fine aggregate criteria and limited fines (3% or less passing the
#200 sieve for Presby systems).

Table 2 – Sieve Analysis for Specified System Sand used on both tests

The fine sand used in Test #1 to simulate native soil is a fine, bank-run sand sourced in
Whitefield, NH.

9
Table 3 – Sieve analysis for Very Fine Sand used in Test #1

The silt used in Test #2 was a waste material sourced from an aggregate company in Lyndon,

VT.
Table 4 – Sieve analysis for silt material used in Test #2

10
RESULTS
Test #1, Specified System Sand over Very Fine Sand. Dosed over the course of two days. Test
ended early due to leakage of water and desire to use less permeable material beneath the
specified System Sand.

Figure 10 – Test #1 – Day 1, Dosing and Plume

11
Figure 11 – Test #1 – Day 2, Dosing and Plume

12
Test #2, Specified System Sand over Silt. Dosed over the course of 5 days. Numbers count
doses.

41

5 6

7 2

3
8

Figure 12 – Test #2 – Day 1, Dosing and Plume

Figure 13 – Test #2 – Day 2 & Day 3, Dosing and Plume; Similar pictures and patterns
for days 4 and 5 not included.

13
DISCUSSION
Both tests in the sand box used the same loading rate, 5 GPD per 1/3 FT of proprietary pipe
(equivalent to 15 GPD/FT of pipe at the test center). When serial distribution is used, this
loading is typical and expected in the beginning, due to the concentration of flows in the first
row and before the biomat has fully developed and expected to be more restrictive to facilitate
treatment (digestion of the organic component of the wastewater). These are actual flows at
test center and design flows which are considered peak flows in the real world.
The sand box tests used clean water to represent the removal of organic waste that would be
the treatment expected for these technologies. However, it should be noted that these flows
would undoubtably be somewhat tempered with the biological treatment and effluent would be
retained for longer periods due to the treatment process itself. A fully mature treatment system
would be expected to use more than one 30-foot row to facilitate treatment. In fact, after 8
months of receiving 450 GPD, it was determined that this particular system used slightly more
than two, 30-foot rows of a 5-row system as evidenced by the visual inspection of the pipes
and conduit immediately after the test was concluded.
For Test #1, this flow rate was readily accepted by the underlying, very fine sand, despite it
being much finer than the system sand. While it was interesting to watch, our goal was to test
an underlying material that couldn’t absorb the water in the immediate area it was being
received and force it to move horizontally for absorption in a larger area.
For Test #2, we used silt as the underlying material. While the silt had a very strong capillary
pull, it was the first material to transmit water to the ends of the box and be fully wetted, it
could not keep up with the volume of water being presented. The sand box itself was able to
absorb and hold onto a significant amount of water. Approximately 10 gallons of water were
being held within the system overall. It took one-and-a-half days for water to begin to be
released from the bottom of the system (noontime on day two). After that, the system quickly
became fully wetted and developed a rhythm and pattern that remained unchanged for rest of
the test. Each dose would release a pulse of water which would leave the proprietary pipe and
travel in a broad shallow pyramid-shaped path (see figure 16), a much broader path than the
capillary force of coarse sand would suggest. It took approximately 5-6 minutes from a dose,
to water leaving the bottom of the system. The pulse would last approximately 30 minutes
from the end of a dose to the end of the pulse with a very visible and quick change in saturation
level in the area of the pipe and below. The water would continue to drain from the bottom of
the box for hours. It is difficult to discern the difference between saturated flow and unsaturated
flow in sand with the naked eye. This was the reason to match a seemingly aggressive loading
rate to a successful secondary treatment testing configuration and actual flow.
An interesting visual indication of the horizontal movement of the water was evident during
test #2 in the unwetted portion of the specified System Sand, which showed a fringe of moisture
along the wetted silt/sand interface in the sand extensions on each side (see figure 15). This is
remarkable because the capillary force of silt is much greater than that of System Sand, as can
be seen in Table 1 of this paper. The only way the sand could be attracting moisture in this
location is for there to be saturated or nearly saturated conditions at the interface, at least in the
locations where this fringe is evident.
These saturated flows or conditions at the sand/silt interface have probably continued to exist,
since the flows to the sand bed were continued at the same volume until the end of the test.
However, it was only visually discernable when the System Sand bed was otherwise dry.

14
 Another interesting occurrence is the development of cracks in the silt layer that was placed
over the System Sand (see figure17). This layer was initially dry and was wetted from below
during the testing and started developing cracks almost immediately during testing when the
level of moisture was relatively consistent. The bottom layer of silt did not develop any cracks
that were evident. Not seeing any implication this would present except to indicate the fine
consistency of this material.

Figure 14 – Drawing of sand box test unit for Test #2 showing initial path of water before
fully wetted.

Figure 16 – Fully wetted sand and silt showing broad dispersal in System Sand.

Figure 15 – Unwetted Specified System Sand indicates path of saturated or nearly saturated
flow at the silt/sand interface due to appearance of a fringe of moisture in the specified sand.

15
 Figure 17 – Silt placed over the System Sand developed large
cracks after initial wetting from below
CONCLUSIONS
Testing a sand box model to demonstrate the process of passive, horizontal movement of
secondary-treated effluent beneath sand-based treatment & dispersal systems is useful to
illustrate the usually unseen parts of the process. Using a liquid loading volume similar to an
actual third-party treatment certification testing that was conducted in serial distribution,
allowed us to see relatively heavy loading of the System Sand bed, knowing similar flows
achieved consistent secondary treatment without the creation of adverse mounding or saturated
conditions, at least in the confines and configuration of the test facility (not real world).
It has been shown through Dr. Gardner’s work that the specified System Sand layer in these
systems would constitute a significant textural change for installations where the horizontal
movement of the effluent would be a necessary criterion (in silty and clay soils for example)
for effective dispersal. Dr. Gardner has shown how water will move around such obstacles,
such as moving laterally along the sand/soil interface. During the sand box test #2, there was
evidence of capillary fringe-like characteristics in the System Sand bed which may be
associated with saturated conditions at the silt/sand interface.
The pulsing of flow observed during the sand box testing showed a quick and marked change
from a heavy flow where a majority of void spaces appeared to be filled with liquid, to a visibly
less saturated state in a relatively short period of time.
Certainly, additional testing is necessary to confirm the results we have seen, build models that
can show further horizontal movement and to further control the variables which may be
influencing the outcomes of our testing. For instance, the polycarbonate sheets we used are
visibly appealing and relatively easy to handle. However, their flexibility has introduced
possible gaps in the cross-section faces that would not exist in real life. This was indicative by
the build-up of condensation on the inside of the panel at times. The worry is that these gaps
would provide a path for saturated water to move vertically via gravity utilizing paths that for
the most part do not exist in the real world.

16
ACKNOWLEDGEMENTS
The author wishes to acknowledge the following people for their help with the sand box
testing and assistance in preparing this manuscript: Lee Rashkin, John Vlk, Dennis Fogg,
Mark Vander-Heyden, Morgyn Stonecipher, Luke Proctor and Kevin Sherman.

REFERENCES

Allen, Dan H., “Hydraulic Mounding of Groundwater Under Axisymmetric Recharge”, Water
Resource Research Center, University of New Hampshire (1980) pg. 20

Amoozegar, A.,Niewoehner, C., Linbo, D. 2006. Lateral Movement of Water in the Capillary
Fringe under Drainfields. Soil Science Department, North Carolina University, Raleigh, NC.
National Onsite Wastewater Recycling Association.

Gardner, Dr. Walter H., “Water Movement in Soils”, USGA Green Section Record,
March/April (1988): 23-27. Print
http://gsrpdf.lib.msu.edu/ticpdf.py?file=/1980s/1988/880323.pdf

Heath, R. C. (1982). “Basic Ground Water Hydrology”, US Geological Survey Water-

Supply Paper 2220, Alexandria VA

Hsieh, J.C.; Gardner, W.H., “Water Movement in Soils”, Washington State University,
(1959). Film https://archive.org/details/educationforlifeadjustment_201512

Sherman, K. (2018), “Capillary Water and its Influence on Common Soils in Florida”.
Florida Journal of Environmental Health 217: 15-19.

Nu, N., Likos, W. (2004) “Unsaturated Soil Mechanics”. John Wiley & Sons, Inc., Hoboken,
NJ

17