32-J28 21
ACTION LEAKAGE RATES FOR LEAK DETECTION SYSTEMS
[Supplemental Background Document for
the Fin? X Double Liners and Leak Detection Systems Rule for
Hazardous Waste Landfills, Waste Piles, and Surface Impoundments]
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Solid Waste
January 1992
U.S Fr*-<•*"Mf*! Protection Agency
' • .- '~'-!2J)
vard, 12th f JOOf
-------
ABSTRACT
This document supplements the background document [Ref. 4]
for the May 29, 1987 proposed double liners and leak detection
systems rule. This supplement explains the application of the
formulas in the original background document to calculate an
action leakage rate (called rapid and extremely large leak in the
proposal), presents the results of action leakage rate
calculations for facilities meeting the minimum design
specifications in the final rule, and provides results from a
more sophisticated 3-dimensional model. The action leakage
rates, based on the minimum specifications in the final rule and
a safety factor of two, are 100 gallons per acre per day (gpad)
for landfills and waste piles, and 1,000 gpad for surface
impoundments. The output from the 3-D model helps to visualize
the shape of the flow for various design specifications and shows
the relative impact of a number of factors on flow capacity.
This supplemental background document also presents
additional data on flow rates actually achieved at a number of
double-lined facilities. These numbers support the proposed and
final rules by showing that facilities with good construction
quality assurance (CQA) perform significantly better than those
without. Further, only about 70% of the well designed facilities
with good CQA meet 20 gpad which was proposed as the upper bound
for a base action leakage rate, and sources of liquids other than
top liner leakage can themselves result in flow rates from the
leak detection system greater than 20 gpad, indicating that the
proposed 20 gpad is too low for a practical action leakage rate.
Finally, this supplemental document also references a number
of technical guidances the Agency has issued since the three
proposals1 that contain useful information relative to all of
the design, performance, monitoring, and response action
standards in the final rule.
1 Proposed in the Federal Register on:
May 29, 1987—Liners and Leak Detection Systems [52 FR.
20218].
March 28, 1986 and April 17, 1987—Double Liners and
Leachate Collection and Removal Systems [51
IB 10706 and 52 IE 12566].
-------
CONTENTS
1 . INTRODUCTION .............................. .> .............. j
2 . ACTUAL FLOW RATES MEASURED IN THE FIELD .................. ]
2 . 1 Data From Commenters ........ . ...................... |
2 . 2 Data From Operating Units .......................... i
2 . 3 Evaluation of Available Information ........... . .... 2.
Landfills with <3eomembrane Top Liners .......... • • • 2-
Surf ape Impoundments with Geomembrane Top Liners . . a.
Landfills with Composite Liners ................... 5-
Surface Impoundments with Composite Top Liners... #
2.4 Theoretical Analysis of Top Liner Performance ...... -7
Available Information ......................
Results of analysis .............. . .........
Frequency and Size of Geomembrane Defects
Analysis Results ........ .
2.5 Summary ...... . ..................................... 0
;... .-y
3 . ACTION LEAKAGE RATE
3 . 1 Determining an Action Leakage Rate ................. \Q
Results Using a 3-D Model
3 . 2 Alternative Action Leakage Rates. ....
3 . 3 Action Leakage Rate Significance
4 . ADDITIONAL GUIDANCES AND REFERENCES
5. CLOGGING ................................
6 . CONCLUSIONS .............................................. (~j
7 . REFERENCES
APPENDICES
A. EPA LINER GUIDANCES
B. FLOW RATE RESULTS USING A 3-D COMPUTERIZED MODEL ......... 2.3
C. BACKGROUND INFORMATION ON THE 3-D VARIABLY SATURATED
FLOW ANALYSIS MODEL ...................................... -
-------
1. TKTRODUCTION
The purpose of this document is to supplement the original
background documents supporting the 1986 and 1987 proposals for
double liners and leachate collection and'removal systems, and
liners and leak detection system rules2 for hazardous waste
landfills, surface impoundments, and waste piles. A lot of
information has been generated since the 1986 and 1987 proposals
that further support this rulemaking. In particular, data on
actual flow rates at double-lined landfills and surface
impoundments and on top liner performance has been collected and
evaluated, flow models have been applied to calculate action
leakage rates, and a number of technical guidances have been-
published. This document discusses each of these.
2. ACTUAL PLOW RATES MEASURED IN THE FIELD
EPA acknowledged in the May 1987 preamble and background
document that it had limited data on the .performance capability
of top liners in -terms of flow rates and stated that the Agency
is seeking additional data. Since the proposal, EPA has gathered
information from a number of facilities, including some data
submitted by commenters. This data is summarized here.
2.1 Data From Commenters
In response to EPA's request for more data, some commenters
(facility operators) submitted actual flow data. One commenter
claimed to achieve, after removal of construction water, a
leakage rate of 2-3 gpad at six landfills and 0, 0, 18, and 75
gpad at four (non-regulated) surface impoundments. Commenters
made a number of claims regarding other sources of liquids in
leak detection systems: consolidation water (from clay in
composite top liners) can be 10-50 gpad; construction water can
be 10-50 gpad; vapor transmission through a top liner geomembrane
can be 4 gpad; and ground water through a geomembrane in the
bottom liner can be 20 gpad.
2.2 Data From Operating Units
Information on top liner performance can be obtained from an
analysis of leachate detection/ collection, and removal systems
(LDCRS) flow rates. The results of field monitoring of LDCRS
flows at double-lined landfills and surface impoundments have
been presented by EPA [1987], Gross et al. [1990], and Bonaparte
and Gross [1990]. The reference by Bonaparte and Gross includes
the data from all the other references cited above, as well as a
2 Proposed in the Federal Register on:
May 29, 1987—Liners and Leak Detection Systems [52 IE
2021U].
March 28, 1986 and April 17, 1987—Double Liners and
Leachate Collection and Removal Systems [51
£R 10706 and 52 £R 12566].
-------
significant body of otherwise unpublished information. The
findings from Bonaparte and Gross are presented in Section 2.3.
2.3 Evaluation cf Available Information
Bonaparte and Gross [1990] presented LDCRS flow rate data
for 55 individually-monitored double-lined landfill cells and 14
individually-monitored double-lined surface impoundments. The
units are located in different climatic regions across the United
States; however, most of the units are located in relatively
moist climatic regions with average annual rainfalls ranging from
35 to 43 in (900 to 1,100 mm). For each unit, they presented
information on the design and operation of the unit, as well as
the rate of flow from the LDCRS. Then they evaluated the
probable sources of the flow from each unit. Potential sources
of flow are illustrated in Figure 1 and include: (i) leakage
through the top liner; (ii) water from precipitation that
percolates into the LDCRS during construction ("construction
water"); (iii) water squeezed out of the clay component of a
composite top liner as a result of clay consolidation
(••consolidation water"); and (iv) ground water that infiltrates
the bottom liner and enters the LDCRS ("infiltration water").
Landfills with Geomembrane Top Liners
In their paper, Bonaparte and Gross evaluated flow rate data
from 23 landfill cells that were constructed with geomembrane top
liners (instead of composite top liners). A geomembrane top
liner represents the minimum technology requirement for top
liners at hazardous waste management units regulated under 40 CFR
Parts 264 and 265. The authors determined that for 16 of the 23
landfills cells, there could be no consolidation water, and,
based on design and operating considerations, construction water
and infiltratin water were unlikely sources of LDCRS flow. As a
result, the measured LDCRS flow could only be attributed to top
liner leakage. Eleven of the 16 landfill cells had been
constructed using construction quality assurance (CQA) procedures
in substantial conformance with EPA [1986] guidance. The other
five cells were constructed using less stringent CQA procedures
or no CQA at all.
Table 1 presents a summary of the data for the 16 landfill
cells constructed with geomembrane top liners. In Table 1, the
LDCRS flow rates are reported in units of gallons per acre of
lined area per day (gpad).
-------
GEOMEMBflANE
GROUND-WATER TABLE
B
TOP LINER LEAKAGE
CONSTRUCTION WATER
CONSOLIDATION WATER
INFILTRATION WATER
LOCRS FLOW
Figure 1. Sources of Flow in Leak Detection, Collection, and
Removal systems (LDCRSs).
-3-
-------
Table 1. Comparison of average LDCR8 flow rates at 16 landfill
cell with geomembrane top liners (from Bonaparte and
Gross [1990]).
No. of Cells
Leakage Detection Laver Plow Rate COA No CQA
Flow rate less than 5 gpad 4
Flow rate in range of 5 to 20 gpad 4 1
Flow rate in range of 20 to 50 gpad 3
Flow rate greater than 50 gpad - 4
From Table 1, it can be seen that of the eleven landfill
cells that were.constructed using a CQA program, four cells had
average flow rates less than 5 gpad (50 liters per hectare per
day (Iphd)), and a total of eight cells had average flow rates
less than 20 gpad (200 Iphd). The data in Table 1 show that a
base leakage rate of 20 gpad (200 Iphd), which is the top of the
range for the base action leakage rate in the proposal, is too
low (i.e., not "practicable") since only 73 percent (eight out of
eleven) of the cells that had properly constructed geomembrane
top liners using rigorous CQA procedures achieved a LDCRS flow
rate of less than 20 gpad (200 Iphd).
Table 1 also provides evidence of the benefit of a rigorous
CQA program. All eleven units constructed using CQA procedures
had LDCRS flow rates of less than 50 gpad (500 Iphd), and eight
of the eleven facilities had flow rates of less than 20 gpad (200
Iphd). In contrast, four of the five units that were constructed
with less rigorous CQA procedures or with no CQA at all had LDCRS
flow rates in excess of 50 gpad (500 Iphd), and two units had
LDCRS flow rates in excess of 100 gpad (1,000 Iphd). At these
two units LDCRS flow rates were on the order of 300 gpad (1,000
Iphd). In summary, the LDCRS flow rates from waste management
units with rigourous CQA programs are significantly lower than
the flow rates from units without rigorous programs.
Based on these data, it appears that flow rates from LDCRSs
of landfills that are properly constructed using rigorous CQA
programs should be well less than 100 gpad (1,000 Iphd). On the
other hand, if a unit is constructed using less rigorous CQA
procedures, LDCRS flow rates greater than 100 gpad (1,000 Iphd)
may occur.
Surface Impoundments with Geomembrane Top Liners
Conclusions similar to those given above for landfills can
also be drawn for surface impoundments. Bonaparte and Gross
[1990] presented data on LDCRS flow rates from eight double-lined
surface impoundments having geomembrane top liners. The authors
determined that for six of these surface impoundments, top liner
leakage was the likely source of any LDCRS flow. Five of the six
surface impoundments were constructed with rigorous CQA programs,
including either ponding tests or leak location surveys; it is
-------
not known if CQA was performed during the construction of the
sixth surface impoundment. The authors reported that four of the
six surface impoundments exhibited no LDCRS flow in the time
period between the start of operation and the time the flow data
was collected. The fifth surface impoundment exhibited no flow
except during a short period between when a geomembrane defect
developed and when it was repaired. The sixth surface
impoundment exhibited a flow of about 0.2 gpad (2 Iphd), except
during a short period when the flow increased to about 40 gpad
(400 Iphd) due to a geomembrane defect. Thus, all six of the
monitored surface impoundments with geomembrane top liners had
LDCRS flow rates below 5 gpad (50 Iphd) except during a short
period between when a geomembrane defect developed and when it
was repaired. This represents an extremely high level of
performance; in fact it represents a higher level of top liner
performance than was observed at landfills .having geomembrane top
liners. This high level of performance was obtained by using
ponding tests and/or leak location surveys as part of the CQA
program. These CQA techniques are typically better adapted to
use at surface impoundments than landfills because surface
impoundments are frequently smaller than landfill cells resulting
in easier implementation of ponding tests or surveying
techniques. In addition, geomembrane top liner defects that may
develop after construction are generally easier to find and
repair in a surface impoundment than in a landfill. The top
liner in a surface impoundment is typically uncovered (or covered
with only a thin veneer of soil), whereas the top liner in a
landfill cell is covered with a drainage layer (leachate
collection and removal system or LCRS) and then with a thick
layer of waste which makes access to the liner difficult.
Based on the available data, it appears that flow rates from
the LDCRSs of surface impoundments that are properly constructed
using rigorous CQA programs (those that use leak location surveys
or ponding tests) should be well less than 100 gpad (1,000 Iphd).
It should be anticipated, however, that if a unit is constructed
using less rigorous CQA procedures, a flow rate greater than 100
gpad (1,000 Iphd) could occur. It is interesting to note that
leak location surveys and ponding tests represent two techniques
that are frequently implemented as part of a response action plan
at surface impoundments experiencing excessive flow from the
LDCRS.. The results described herein suggest that these
techniques will be effective in reducing the LDCRS flow rate to
below 100 gpad (1,000 Iphd) at surface impoundments for which
response actions are required.
Landfills with composite Top Liners
The evaluations discussed above were for double-lined units
having geomembrane top liners. It is also useful to consider
units having composite top liners to assess the contribution of
consolidation water from the clay component of the top liner
toward potentially exceeding an action leakage rate. Although
the action leakage rate in the final rule, as in the proposal, is
based on total flow in the LDCRS, regardless of source, the
response actions should consider sources other than leaks. For
-------
this reason, it is relevant to compare LDCRS flow rate data from
units with composite top liners.
Bonaparte and Gross [1990] evaluated LDCRS flow rate data
from 32 landfill units with a composite top liner. Because the
top liner is a composite liner, the primary source of the flow
can be attributed to construction water plus consolidation water
if an analysis of the time required for leakage to flow through
the top liner (i.e., leakage breakthrough time) was greater than
the time since the end of construction of the landfill. For 18
of these units, the authors attributed the flow primarily to
construction plus consolidation water from the clay component of
the composite top liner. Data on these 18 units are provided in
Table 2.
Thirteen of the waste management units used to generate the
data in Table 2 were constructed using CQA programs in
substantial accordance with EPA [1986] guidance; four were
constructed without CQA programs; and it is not known if CQA was
performed during construction of the remaining unit.
v
Table 2. Average LDCRS flow rates at 18 landfill cells with
composite top liners (from Bonaparte and Gross [1990]).
Leak Detection Layer Flow Rate No. of Cells
< 5 gpad 5
5 to 20 gpad 8
> 20 to 50 gpad 3
> 50 to 100 gpad 2
> 100 gpad 0
From Table 2, it can be seen that only five of 18 (28
percent) of the landfill cells constructed with composite top
liners have LDCRS flow rates of less than 5 gpad (50 Iphd), and
13 of 18 cells (72 percent) have LDCRS flow rates of less than 20
gpad (200 gpad). This data is similar to that for geomembrane
only top liners, indicating that construction water is rather
insignificant at these units (perhaps because overburden
pressures have yet to squeeze out the consolidation water). This
data also indicates that perhaps a significant source of the
liquids at the geomembrane only units is construction water. At
any rate, this data further supports the conclusion that an
action leakage rate of 20 gpad (200 Iphd) is inappropriate since
it would mean most waste management units with composite top
liners will also have LDCRS flow rates that exceed the action
leakage rate under normal operating conditions.
All 18 units with composite top liners exhibited average
LDCRS flows below 100 gpad (1,000 Iphd). Thus, it appears that
properly constructed waste management units with composite top
-------
liners are unlikely to exhibit LDCRS flows that exceed 100 gpad
(1,000 Iphd).
Surface Impoundments vitb Composite Top Liner*
There is insufficient data to present observations on the
performance of this category of facilities. However, it is
anticipated that the performance of these facilities would be the
same as the performance of landfills with composite top liners.
2.4. Theoretical Analysis of TOP Liner Performance
A theoretical analysis of top liner performance was also
performed. This analysis further supports the conclusion from
the above data that 20 gpad is not a practical action leakage
rate.
Available Information
In recent years, various investigators have developed
analytical techniques for estimating leakage rates through
liners. These investigations include: Bonaparte et al. [1989];
Brown et al. [1987]; EPA [1987]; Giroud and Bonaparte [1989a,b];
Giroud et al. [1991]; and Jayawickrama et al. [1987]. The
reference presented by Bonaparte et al. [1989] presents equations
to estimate leakage rates through both geomembrane liners and
composite liners; these equations are used in the analysis below
to estimate leakage rates through top liners.
To estimate the anticipated leakage rate through a top liner
at a waste management unit, a frequency of defect and size of
defect in the geomembrane component of the top liner must be
assumed. Available information on the frequency and size of
defects in properly-installed geomembrane liners had been
reported by EPA [1987], Giroud and Bonaparte [1989a], Giroud and
Fluent [1987], and Laine [1991]. This information is also used
below to estimate leakage rates through top liners.
Results of Analysis
Frequency and Size of Geomembrane Defects. Giroud and Bonaparte
[1989a] presented limited case study data, including CQA records,
records of foresnic investigations, and LDCRS flow rate data,
from which they drew "tentative" conclusions regarding the
frequency and size of defects in geomembrane liners installed
using rigorous CQA procedures. From their data, they recommended
that for the purpose of estimating leakage rates through
geomembranes, a geomembrane defect (hole) frequency of one to two
per acre (two to five per hectare) be considered along with a
defect size of 0.005 in2 (3.2 mm2). Recently Laine [1991]
presented data from two leak location surveys in which
geomembrane seam defects were identified at a frequency of two to
five per acre (five to twelve per hectare). Thus, the frequency
of defects found by Laine is twice as high as the frequency
recommended by Giroud and Bonaparte for estimating leakage rates.
However, the size of the defect found by Laine was typically very
-------
small, i.e., pinhole sized with areas on the order of 0.001 in2
(0.6 mm2) or less. The defect size is about five times smaller
than the defect size recommended by Giroud and Bonaparte for
estimating leakage rates. Since the calculated leakage rate for
a given installed area of geomembrane is proportional to the
product of the size of the defect and the frequency of defects,
the findings of both of the above-described investigations lead
to comparable top liner leakage rates when used.
For the analysis of top liner leakage rates presented below,
a defect frequency of one per acre (two per hectare) and a defect
size of 0.005 in2 (3.2 mm2) is assumed.
Analysis Results. The results of calculations using the
equations from Bonaparte et al. [1989] for steady-state leakage
through geomembrane holes are presented below. For the
calculations, it was assumed that the top liner consists of a
geomembrane alone, and the hydraulic conductivity of the material
overlying the geomembrane is 1 x lp"2 cm/s (1 x 10"* m/s) which is
appropriate for a landfill with a granular,leachate collection
and removal system (LCRS). The calculated top liner leakage
rates, given the above-described conditions, are presented in
Table 3.
Table 3. Calculated leakage rates through a geomembrane top
liner.
Liquid head on Steady-State
top liner - leakage rate
0.1 10
1.0 60
10.0 220
Calculated top liner leakage rates would be much lower than
those given in Table 3 if the top liner was a composite liner
rather than a geomembrane alone. Conversely, the calculated top
liner leakage rate would be somewhat higher if the material above
the top liner had a higher permeability, or if the liner was
exposed (as might be the case for a surface impoundment) .
The calculation results presented above must be interpreted
separately with respect to landfills and surface impoundments.
For landfills, the design maximum liquid head in the LCRS is 1 ft
(0.3 m) . However, the average liquid head under normal operating
conditions should be only on the order of 0.1 ft (0.03 m) ; in
many instances, the average head may be only on the order of 0.1
ft (0.03 m) , or even less. In this case the calculated results
support a conclusion that under normal operating conditions
(i.e., when there is an average hydraulic head in the LCRS of 0.1
ft (0.03 m) , or less), the leakage rate through a properly
designed geomembrane top liner, constructed using proper
procedures and rigorous CQA, will frequently be less than 20 gpad
-------
(200 1 phd). During periods of maximum leachate flow (e.g,
after major storm events), top liner leakage rates in landfills
with geomembrane top liners could temporarily exceed 20 gpad (200
Iphd) and approach 60 gpad (600 Iphd), since the liquid head in
the LCRS during this period could easily exceed 0.1 ft (0.03m).
The calculation results suggest that for surface
impoundments constructed with geomembrane top liners (where the
liquid head may be on the order of 10 ft (3 m)), top liner
leakage rates could easily exceed 20 gpad (200 Iphd) and approach
200 gpad (2,000 Iphd) even if there is only one small geomembrane
defect per acre (two defects per hectare) of liner. Thus, to
keep top liner leakage rates below 20 gpad, or even 200 gpad, in
surface impoundments with geomembrane top liners, geomembrane
defects need to be virtually eliminated. In most cases, this
will only be accomplished usi.ng ponding tests, leak location
surveys, or other "extraordinary" CQA procedures. As shown by
the monitoring data presented in Section 2.3 of this report, when
these CQA procedures are used, top liner leakage can be largely
eliminated, at least for.some period of time.
v
2.5 Summary
As stated in the proposal, and restated by some of the
commenters, the existing empirical data base at the time of the
proposal regarding actual flow.rates was quite limited. EPA has,
however, accumulated empirical data since the proposal on the
performance of different liner designs. This data help give
meaning to different flow rates in terms of the ability of
owner/operators and technology to achieve and in terms of leaks
versus other sources of liquids. This additional leakage rate
data are consistent with the data submitted by the commenters.
The actual flow rate data presented above are summarized in
Table 4, for all 40 units.
Table 4. Actual Flow Rates at Double-Lined Individually-
Monitored Landfill and surface Impoundment Units.
LDS FLOW RATE (GPAD)
< 5
5-20
>20-50
>50
NO. of UNITS
15
13
6
6
% of UNITS
38
32
15
15
NOTES TO TABLE: These are units where other sources,
except construction water, were determined not to be a
factor. Thirty-one of the 40 units were constructed
with rigorous CQA, 7 were not, and 2 are unknown. Of
the six at >50 gpad, at least four had no rigorous CQA.
This data shows that only 70% of the 40 units meet 20 gpad; and
only 85% of the 40 units, but at least 95% of the units with
rigorous construction quality assurance (CQA), meet 50 gpad.
This indicates that 20 gpad and even 50 gpad are not practicable
action leakage rates for the general situation.
-9-
-------
This data in conjunction with the previous EPA data show that
over the past 10 years, and especially in more recent years,
facility owners and operators have been building and operating
liner systems that work better and better to minimize flow
through the top liner. The major contributions to this
improvement have been better installation practices and better
CQA.
3. ACTION LEAKAGE RATE
In the final rule, as in the May 29, 1987 proposal, the
owner or operator of units subject to the leak detection system
requirements must propose and the Regional Administrator (or
State Director in authorized states) must approve an action
leakage rate. "Action leakage rate" is defined in the final- rule
as "the maximum design flow rate that the leak detection system
(LDS) can remove without the fluid head on the bottom liner
exceeding 1 foot. The action leakage rate must include an
adequate safety margin to allow for uncertainties in the design
(e.g., slope, hydraulic conductivity, thickness of drainage
material), construction, operation, and location of the LDS,
waste and leachate characteristics, likelihood and amounts of
other sources of liquids in the LDS, and proposed response
actions (e.g., the action leakage rate must consider decreases in
the flow capacity of the system over time resulting from
siltation and clogging, rib layover and creep of synthetic
components of the system, overburden pressures, etc.)." In
short, the "action leakage rate" is the maximum design flow rate,
with a safety factor, that the leak detection system can remove
without the head on the bottom liner exceeding one foot (called
rapid and extremely large leak in the May 29, 1987 proposal).
The objective is to minimize the head or pressure on the bottom
liner and thereby decrease the potential for migration of
hazardous constituents out of the unit should a leak in the
bottom liner, as well as the top liner, occur. The proposal
background document [Ref. 4] presented a number of mathematical
models for making such a determination. All of these models are
based on Darcy's Law for non-turbulent flow through saturated
media.
3.1 Determining an Action Leakage Rate
The proposal background document gives the following formula
for flow originating through a hole in the liner, the most likely
leak scenario for a geomembrane liner (pages 2.6-12 and 2.10-10,
Ref. 4):
Q s k-h«tan ct'B^ [Equation 1]
where Q « flow rate in the leak detection system
(drainage layer),
h * head on the bottom liner,
k « hydraulic conductivity of the drainage
medium,
a - slope of the leak detection system,
-------
B,
•vg
average width of the flow in the leak
detection system, perpendicular to the flow.
Assumming that the gradient of flow through the hole, at the
hole, is sin a and depth of flow at the hole for concentrated
flow « the thickness of the drainage layer:
where
B«vg " D/sin o
leak detection system thickness.
Then, with D - 1 ft and sin a - 0.01, B - 100 ft
0.02, *Fl - 50 ft
0.03, B^J - 33 ft.
Using these values for B and Equation 1 with h & D « 1 ft
D for small values of o), Q in gpad «
k
(cm/ sec)
1
.1
.01
sin a
.01
.02
.03
.01
.02
.03
.01
.02
.03
B.V9 (ft)
33
....
....
21,000
....
2,100
....
210
50
....
21,000
....
....
2,100
r "*""""
....
210
^^«MM»
100
21,000
— _
• M«W
2,100
....
210
....
^»^^ «M
Thus, using the minimum specifications in today's rule: 1% slope,
12 in thick drainage layer, and l X 10*1 cm/sec hydraulic
conductivity for surface impoundments and 1 X 10*2 cm/sec
hydraulic conductivity for landfills and waste piles, and
assuming that the head is 1 ft and the average width of flow
(B ) is as given above, the results show maximum flow rates of
2,100 gpad for surface impoundments and 210 gpad for landfills
and waste piles. Using a safety factor of two, as suggested in
the example given in the proposed rule preamble, yields about
1,000 gpad for surface impoundments and 100 gpad for landfills
and waste piles as the Agency recommended action leakage rates,
for units that are designed to the minimum specifications in
today's rule. As listed in the rule and above, the safety factor
helps account for uncertainties in the design, construction,
operation, and location of the drainage layer and potential
decreases in flow over time as a result of overburden compressive
forces and clogging caused by fines and biological and chemical
actions in any leachate that seeps through. Of course, all of
the above mechanisms that could result in potential decreases in
flow over time should also be considered when selecting the
design, especially the hydraulic conductivity of the drainage
layer, and in construction. Because this calculation used the
-------
minimum technical requirements and other design assumptions to
maximize potential head on the bottom liner, and uses a safety
factor, EPA believes that the units meeting the minimum technical
requirements would not require action leakage rates below 100
gpad for landfills and waste piles and 1,000 gpad for surface
impoundments.
Assuming the wetted area in the drainage layer beneath a small
hole leak has approximately the shape of a cone from side-view
and a parabola from top. view, the width of the parabola (B) is:
B -
where x - plan distance downslope from hole (i.e., B is a
function of the distance x from the hole; most of B is
at the hole with only slight increases downslope).
Assuming x - 0 (i.e., looking at B under the hole, B - ^u»«l" )
and substituting this value for B into Equation 1 modified for a
triangular cross-section of flow (i.e., Q » 1/2 k»h»tan a«B) and
solving for Q yields:
Q = k'h2 [Equation 2]
where h « head on the bottom liner and h < thickness of
drainage layer.
This equation becomes the following if the condition is changed
from "h < thickness of the drainage layer (D)" to "h £ D" (which
is important for geonet calculations):
Q s k»D (2h - D) [Equation 3].
Solving Equation 3 using the minimum design specifications in the
final rule, Q -
for . 1 cm/sec: 2100 gpad
,01 cm/sec: 210 gpad
geonet: 6800 gpad.
These numbers are the same as the results given above for
Equation 1.
Results Using a 3-D Model
Tables 1-4 and Figures 1-10 in Appendix B were developed from a
3-D model to show the relative effects of various design
parameters and assumptions on flow capacity, and to show the
shapes of the flow in the drainage layer for various designs and
assumptions, including hole size and head. Appendix C gives
background information on the 3-D model. The tables show thc.t
slope, length of run, and hole size have some effect on flow rate
(e.g., 4% increase in flow rate when slope is increased from 1%
to 2% [Tables 1, 3-5]; 1% increase in flow rate at 1% slope when
-12.-
-------
increasing length of run from 20 ft to 80 ft [Table 1; Figure 4
shows that length of run has negligible effect for slopes at or
greater than the 1% minimum]; 43% increase when hole size is
increased from .25 ft2 to 1.0 ft2 but a much less significant
increase for holes > 3 ft2 [Table 2; Figure 5 graphically shows
the effect of leak size on flow rates]). However, the effect of
these three variables is relatively insignificant compared to
hydraulic conductivity, head, and drainage layer thickness (e.g.,
ten times increase (900%) when increased from .01 cm/sea to .1
cm/sec hydraulic conductivity [Tables 1, 3-5]; 382% increase when
increased from no head to 2 ft head above the top liner, e.g., in
a 2 ft deep surface impoundment [Table 3]; and 210% increase when
geonet thickness is doubled from 5 mm to 10 mm [Table 5]).
Figures 2a-2d (side view) and 3a-b (top view) show the shape
of the saturated zone for various designs, assuming no head above
the top liner. These show only small portions of the bottom
liner are actually exposed to the 1 ft head (as assumed in the
simpler models discussed above). Figures 6-8b, however, show
that as the head increases, so does the area of the bottom liner
exposed to the greater heads. The graph for 8 ft head for
surface impoundments is almost rectangular and therefore is not
shown. Table 5 and Figure 10 show the results for geonets, which
because of their high hydraulic conductivities have high flow
rates.
Table 4 shows flow rates of 204 gpad and 2,040 gpad
respectively for the landfill and surface impoundment
specifications (i.e., 1% slope and hydraulic conductivity of 10"1
cm/sec for surface impoundments and 10"1 cm/sec for landfills,
but with 1 ft of head above the top liner, 180 ft length of run,
and a 1 ft2 hole size). Comparing the results of the 3-D model
to those of Equations 1 and 3, using the 1% slope and 10"1 cm/sec
hydraulic conductivity for surface impoundments, shows that if
the hole size is somewhat less than .25 ft2' the flow rate with a
2 ft head would be about 2100 gpad [Table 3]. For 0 ft head
above the top liner, the hole would be somewhat larger than 30
ft2, or close to uniform flow [Figure 5].
3.2 Alternative Action Leakage Rates
While EPA recommends the above action leakage rates (100 and
1,000 gpad) for units that are built to the minimum design
specifications, the Agency recognizes that a number of site-
specific factors affect the maximum flow capacity of a leak
detection system, and owners and operators may want to propose
alternative action leakage rates. For example, the leak
detection system design may be different than the minimums
specified in the final rule. As indicated above, the hydraulic
conductivity is a factor that significantly affects the flow
capacity of the system. Since they are directly proportional, a
ten times increase in hydraulic conductivity (i.e., from 10"2 to
10"1 cm/sec) increases the flow capacity ten times. Therefore,
EPA believes that leak detection systems with greater hydraulic
conductivities would have higher action leakage rates. In
addition, owners or operators may have information to justify a
-------
different width of flow in the above calculation. Or the owners
or operators may justify a higher action leakage rate by using a
different formula or model. While the Agency recommends the use
of the above model for defining the maximum flow capacity of the
leak detection system and action leakage fate, EPA recognizes
that there may be alternative models available now or in the
future that may more accurately predict system flow capacity to
justify higher action leakage rates. Therefore owners or
operators may propose to use an alternative model that they
believe more accurately predicts the maximum flow capacity of the
leak detection system. Or, owners or operators may want to do a
field flow (pump) test on the leak detection system to show
actual flow capacity, which may justify a higher action leakage
rate. Finally, owners or operators may have flow rate data on
similarly designed units to use to justify a different level. As
more and more units are built, the Agency as well as owners .or
operators will develop a better data base that may be used to
justify other action leakage rates.
3.3 Action Leakage Rate significance
Action leakage rates must not exceed the maximum flow rate
capacity of the leak detection system in order to assure that a
response action is triggered for significant leaks. That is, if
the action leakage rate were greater than the flow capacity of
the system, the trigger level or action leakage rate would never
be reached and response actions implemented, no matter how large
or massive the failure. Further, an action leakage rate that is
based on a maximum of 1 ft head assures that significant
pressures on the bottom liner will not be experienced, thereby
decreasing the potential for migration of hazardous constituents
into the bottom liner. Finally, EPA believes that flow rates in
excess of the minimum action leakage rates often indicate a major
localized or general failure of the top liner. Flow rates of
1,000 gpad or greater represent significant flow rates and
potentially significant hole sizes that may be readily identified
and repaired. Flow rates between 100 gpad and 1,000 gpad are i/
large enough that the sources other than a leak will probably not '~/f\
account for all the flow (i.e., there is probably a leak
situation that should be looked into). For these reasons, it is
necessary to maintain leak detection flow rates below the action
leakage rate and for the owner or operator to take response
actions for leaks greater than the action leakage rate.
The appropriate response action must be based on site-
specific circumstances, including the magnitude of the actual
flow rate (which is related to leak size), the ease of
determining the source of leak and repairing it (e.g., often in a
surface impoundment a hole can be observed from the surface, or a
bulge in the top liner from underlying pressures may be observed
from the surface indicating the possible leak location), and
status of the unit (e.g., for a disposal unit about to close, it
may be best to close the unit and get a sound cover on top rather
than seek to find and repair a leak, especially for relatively
low flow rates).
-/¥-
-------
4. ADDITIONAL GUIDANCES AND REFERENCES
A number of technical guidance manuals have been published
by EPA that discuss all the design features of the final rule.
Some of these are listed in Appendix A. These cover: foundations
and dikes; flexible membrane liners or geomembranes; soil/clay
liners; composite liners; hydraulic conductivity and other
properties of granular drainage layers, geonets, and clay/soil
liners; leachate collection, and removal systems and leak
detection systems designs; sumps and pumps; clogging;
construction quality assurance and test fills; Darcy's Law and
calculation of flow quantities, flow capacities, and time of
travel or breakthrough times; response action plans; and covers.
5. CLOGGING
EPA sponsored studies [Bass et al., 1183; Bass, 1986; Ghasseml et
al., 1986; Koerner et al.v 1991] Indicate that clogging of drainage
layeVs of waste management units may potentially occur under some
conditions. The results of the studies Indicate that drainage layer
clogging Is caused primarily by sedimentation or biological growth. The
results of the studies also suggest that the potential for clogging can
be minimized by proper design and'construction of the drainage layer.
The potential for clogging of LDCRSs is generally lower than that for
overlying leachate collection and removal systems (LCRSs) due to the
relatively low volumes of flow in LDCRSs. Clogging of LDCRSs, however,
could hinder the detection of leakage and the rapid removal of liquid
from the LDCRS.
With this in mind, EPA is supporting the use of relatively permeable
LDCRS materials in waste management units to minimize their potential for
clogging. Fundamentally, a drainage material with large particles and,
hence, large pore spaces, would have less potential for clogging than a
material with smaller particles and, hence, smaller pore spaces. That
is to say that materials such as coarse sands and fine gravels with a
minimum hydraulic conductivity of 1 cm/s (1 x 10~2 m/s) would be less
likely to clog than materials such as fine sand or silty sand with a
minimum hydraulic conductivity of 1 x 10"2 cm/s (1 x 10~4 m/s).
-iff-
-------
Landfill Clogging1
Following is a summary of a research study looking at
clogging.
Tested: * 2x10-2 cm/sec Ottawa sand (subrounded uniform size,
0.42mm—no. 40 sieve—avg particle size);
* 5x10-3 to 4.7x10-1 cm/sec filter fabrics (7 different
geotextiles, including polypropylene (PP), polyethylene
(PE), and polyester (PET)).
[Note: geotextiles use less space, are easier to transport,
easier to place, and less expensive].
Tested using municipal waste leachate of different strengths.
Conclusions
* Flow rates always decreased (from 10-100%) over time: usually
a sharp initial decrease followed by a continued linear,
slightly linear, or sharply exponential decrease. In some
cases flow decreased to levels that were not measurable by the
experimental design.
* Sand (over geotextiles) clogged considerably more than those
with geotextiles alone (23% flow retained for sand/geotextiles
vs 34-45% flow retained with geotextiles alone).
* Type of polymer (PP, PE, & PET) appears to have no
significance. Biological degradation of polymeric-based
geotextiles did not occur.
* Stronger leachates (i.e., with higher BOD, COD, & TS) have
greater clogging impacts. Particulate clogging appeared to be
synergistic with the biological clogging.
* Both anaerobic and aerobic conditions promote clogging.
Robert and George Koerner, Landfill
600/2-91/025, August
1991 (NTIS * PB91-213660)
-------
6. CONCLUSIONS
Facilities with good CQA perform significantly better than those
without.
Facilities with good CQA appear to consistently achieve 50 gpad
or less, taking into account other sources of liquids such as
construction water and consolidation water. Whereas only about
70% of the facilities with good CQA achieve 20 gpad, which was
the top of the range in the May 29, 1987 proposed rule. These
results coupled with the magnitude of other sources of liquids
indicates a practical action leakage rate is £ 100 gpad.
Calculations and models used to determine the action leakage rate
show:
* Flow rates of 100 gpad for landfills and waste piles and
1,000 gpad for surface impoundments appear to be reasonable
action leakage rates for the minimum specifications for
slope and hydraulic conductivity in the final rule;
* Hydraulic conductivity is a significant factor (in all the
models) since the flow rate is directly proportional to
hydraulic conductivity: a change from 10"2 to 10"1 cm/sec
increases the flow rate 10 times;
* Slope is relatively insignificant;
* Length of run is not a factor for slopes £ 1%;
* With no head above the top liner, the shape of flow is
basically conical below the hole and rapidly tapers off, but
with heads above the top liner more of the bottom liner is
exposed to the higher heads;
* The size of leak is a factor that also influences whether
the action leakage rate or flow capacity of the leak
detection system will be exceeded. In the formula in the
proposal background document, the size of leak is not
considered since it is assumed that the hole is large enough
to provide the maximum flow rate (Q) calculated. The 3-D
model however confirms that the size of leak is indeed a
limiting factor;
* Models that assume uniform leakage (which is an unrealistic
assumption because the top liner is a geomembrane, not clay
or other porous media) give higher flow capacities than
'models assuming one or more leaks through the top liner.
Clogging by fines or biological and chemical actions needs to be
considered in the design (e.g., by the use of gradation or fabric
filters and higher permeability drainage materials) and in the
safety factor.
-------
7. REFERENCES
1. Bonaparte, R., Giroud, J.P., and Gross, 8.A., "Rate of Leakage
Through Landfill Liners". Proceedings, Geosynthetics '89,
San Diego, California, Feb 1989, Vol.- 1, pp. 18-29.
2. Bonaparte, R. and Gross, B.A., "Field Behavior of Double-Liner
Systems". Proceedings. Waste Containment Systems! Construc-
tion. Regulation, and Performance. San Francisco,
California, ASCE Geotechnical Special Publication No; 26,
Nov. 1990, pp. 52-83.
3. Brown, K.W., and J.C. Thomas, "Leak Rates into Drainage
Systems Underlying Lined Retention Facilities", Journal of
Hazardous Materials, 18 (1988) p.179-188.
4. Brown, K.W., Thomas, J.C.r Lytton, R.L., Jayawickrama, P., and
Bahrt, S.C., "Quantification of Leak Rates Through Holes in
Landfill Liners". USEPA Report CR 810940, Cincinnati, Ohio,
1987, 147 p.
5. EPA, "Background Document! Proposed Liner and Leak Detection
Rule". EPA/53O-SW-87-015, May 1987, 526 p.
6. EPA, "Landfill Leachate Clogging of Geotextile (and Soil)
Filters". Geosynthetic Research Institute, Drexel University
for EPA, CR-814965.
7. EPA, "Technical Guidance Document: Construction Quality
Assurance for Hazardous Waste Land Disposal Facilities".
EPA/530-SW-86-031, Oct. 1986, 88 p.
8. Giroud, J.P., Badu-Tweneboah, K., and Bonaparte, R., "Rates of
Leakage Through Composite Liners Due to Geomembrane
Defects", accepted for publication in Geotextiles and
Geomembranes. 1991.
9. Giroud, J.P. and Bonaparte, R., "Leakage Through Liners
Constructed with Geomembranes - Part I. Geomembrane
Liners", Geotextiles and Geomembranes. Vol. 8, No. 1, 1989a,
pp. 27-67.
10. Giroud, J.P. and Bonaparte, R., "Leakage Through Liners
Constructed with Geomembranes - Part II. Composite Liners",
Geotextiles and Geomembranes. Vol. 8, No. 2, 1989b, pp. 78-
111.
11. Giroud, J.P., Bonaparte, R., Ah-Line C., and Beech, J.F.,
"Design of Leakage Detection Layers in Double-Liner
Systems", submitted for publication to Geotextiles and
Geomembranes.
12. Giroud, J.P. and Fluet, J.E., "Quality Assurance of
Geosynthetic Lining Systems", Geotextiles and Geomembranes.
Vol. 3, No. 4, 1986, pp. 244-287.
-------
13. Gross, B.A., Bonaparte, R., and Giroud, J.P., "Evaluation of
Flow From Landfill Leakage Detection Layers", Proceedinas.
Fourth International Conference on Geotextiles. Vol. 2, The
Hague, Jun 1990, pp. 481-486.
14. Jayawickrama, P., Brown, K.W., Thomas, J.C., and Lytton,
R.L., "Leakage Rates Through Flaws in Geomembrane Liners",
Journal of Environmental Engineering. ASCE, Vol. 14, No. 6,
Dec 1988, pp. 1401-1420.
15. Laine, D.L., "Analysis of Pinhole Seam Leaks Located in
Geomembrane Liners Using the Electrical Leak Location
Method: Case Histories", Proceedings. Geosvnthetics '91.
Vol. 1, Atlanta, Feb 1991, pp. 239-253.
-II-
-------
APPENDIX A
EPA LINER GUIDANCES ,
DESIGN t CONSTRUCTION OF LINER SYSTEMS
Al. Guide to Technical Resources for the Design of Land Disposal
Facilities. EPA-625-6-88-018, December 1988, 63p.
A2. Seminars—Requirements for Hatardous Waste Landfill Design.
Construction and Closure. CERI-88-33, June 1988, 441p.
A3. Seminar Publication! Requirements for Hazardous Waste
Landfill Design, construction, and Closure. EPA-625-4-89-022,
CERI, August 1989, 127p.
A4.
EPA-600-2-88-052, RREL, Sept. 1988, 1026p.
for Hazardous Waste Land Disposal Facilities. EPA-530-SW-86-031,
Oct. 1986, 99p.
CLAY/SOIL LINERS
A6. Design* Construction, and Evaluation of Clay Liners for Waste
Management Facilities. EPA-530-SW-86-007F, Nov. 1988.
FML SEAMS
A7. Technical Guidance Document; The Fabrication of Polyethylene
FML Field Seams. EPA-530-SW-89-069, Sept. 1989, 42p.
A8. MEMO: "Use of Construction Quality Assurance (CQA) Programs
and control of Stress Cracking in Flexible Membrane Liner Seams",
Sylvia Lowrance to HWMDDs, Regions I-X, July 13, 1989, 14p.
A9. Field Inspector'a Manual: Stress Cracking of Flexible
Liner Seams. EPA, December 1988.
COVERS
A10. Design and Construction of Covers for Solid Waste Landfills.
EPA-600-2-79-165, MERL, August 1979, 274p.
All. Technical Guidance Documentt Final Covers on Hazardous Waste
Landfills and Surface Impoundments. EPA-530-SW-89-047, July 1989,
39p.
-20-
-------
LINER PERFORMANCE REFERENCES
GENERAL LINER FAILURE ANALYSES
A12. Expected Life of Synthetic Liners and Caps (Draft Final
Report). W. Lyman et al, Arthur D. Little, Inc. for EPA, March
31, 1983.
A13. Performance of Clav Caps and Liners for Disposal Facilities
(Final Report). Research Triangle Institute for EPA, March 1983.
A14. Clay Cap and Liner Systems (Draft Final Interim Report 11f
The Earth Technology Corp. (Ertec) for EPA, June 1983.
A15. Land Disposal Liner/Locational Analysis Project fRevised
Draft Final1. The Earth Technology Corp. for EPA, January 1984.
A16. Evaluation of Flexible Membrane Liner Seams and Evaluation
of Flexible Membrane Liner Systems After Exposure and Simulated
Weathering [same report?), W. Morrison and L. Parkhill, US Bureau
of Reclamation for EPA, c. December 1985.
A17. Background Document on Bottom Liner Performance in Double-
Lined Landfills and Surface Impoundments. GeoServices Inc.
Consulting Engineers, April 1987.
A18. Background Document on Proposed Liner and Leak Detection
Rule. GeoServices Inc. Consulting Engineers for EPA, May 1987.
A19. Performance Analysis of Alternative and Minimum Technology
Designs for Landfills. Surface Impoundments, and Waste Piles.
Radian Corp. for EPA, August 1987.
A20. Field Behavior of Double-Liner Systems. Rudolph Bonaparte
and Beth Gross, Waste Containment Systems: Proceedings of ASCE
Symposium, SFO, Nov. 6-7, 1990.
A21. Quantification of Leak Rates Through Holes in Landfill
Liners. K. Brown et al for EPA, August 1987.
A22. Draft Background Document on Double Liner Rule. EMCON
Associates for EPA, September 1987.
A23. "Durability and Aging of Geosynthetics—2nd GRI Seminar",
December 8 & 9, 1988, 21 papers.
Liner-Waste Compatibility
A24. Liner Materials Exposed to Hazardous and Toxic Wastes. H.
Haxo, Jr. et al, Matrecon, Inc. for EPA, September 1984, 271 pgs.
A25. "Liner Materials Exposed to Hazardous and Toxic Wastes",
Waste Management & Research (1986) 4, 247-264, H. Haxo, Jr. et
al.
-2.1-
-------
A26. "Analysis and Fingerprinting of Unexposed and Exposed
Polymeric Membrane Liners", H. Haxo, Jr., Matrecon, Inc.
A27. "Supplementary Guidance on Determining Liner/Leachate
Collection System Compatibility" memo from Bruce Weddle to EPA
Regions, August 7, 1986.
A28. Analysis of Flexible Membrane Liner Chemical Compatibility
Tests (Draft Final Report). A. Schwope et al, Arthur D. Little,
Inc. for EPA, March 31, 1983.
FML Stress cracking
A29 * Environmental Stress Cracking of HOPE Geomembrane Seams and
Related Studies. Geosynthetic Research Institute for EPA, March
20, 1988.
Permeability/Hydraulic Conductivity cf Clays/Soils
See A6 and Al-5.
A30. Procedures for Modeling Flow Through Clav Liners to
Determine Required Liner Thickness. EPA/530-SW-84-001, April
1984.
A31. Soil Properties. Classification, and Hydraulic Conductivity
Testing. SW-925, March 1984.
"Corrective" Technologies
A32. Assessment of Technology for Constructing and Installing
Cover and Bottom Liner Systems for Hazardous Waste Facilities.
Vol. I: Data Base Development; Perspectives of Industry Experts.
State Regulators, and Owners and Operators; Vol. II; Technical
Analysis (Final Report1. TRW for EPA, April 1983.
-------
APPENDIX B
FLOW RATE RESULTS USING A 3-D COMPUTERIZED MODEL
-------
Leakage from Top Liner and Flux Through
Drainage Layer for Double Lined Landfills and
Surface Impoundments: Computer Simulations
For
Liner and Leak Detection Rule
January 1992
Technical Assessment Branch
Characterization and Assessment Division
Office of Solid Waste
Washington, D. C. 20460
-------
CONTENTS
1. Contents i
2. Acknowledgements ii
3. Introduction 1
4. Modeling Approach l
5. Model Input Data 2
6. Model Output * 2
7. Modeling Assumptions 2
8. Simulation Scenarios 3
9. Simulation Results 3
10. List of Tables 4
11. Tables 1 through 5 5
12. List of Figures 11
13. Figures 1 through 10 12
-------
This report was prepared by Dr. Zubair Saleem, Office of
Solid Haste, U.S. Environmental Protection Agency. The model
runs were performed using a three-dimensional finite element
model, VAM3D - CG (Variably Saturated Analysis Model in Three
Dimensions - Conjugate Gradient). The model employs a
preconditioned conjugate gradient matrix solution scheme which
allows several thousand nodal unknowns to be solved efficiently.
The model runs reported in this report were performed at
HydroGeoLogic, Inc. (HGL), by Dr. Sorab Panday, Dr. Namsik Park,
Mr. John Doyle, and Mr. Amit Sinha. Dr. Ed Sudicky of University
of Waterloo, Dr. Peter Huyakorn and Jack Robertson, of HGL, and
Dr. Michael Ungs of McLaren/Hart provided helpful suggestions.
-------
LEAKAGE FROM TOP LINER AND FLUX THROUGH DRAINAGE
LAYER OF DOUBLE LINED LANDFILLS AND SURFACE
IMPOUNDMENTS: COMPUTER SIMULATIONS
INTRODUCTION
The Hazardous and Solid Waste Amendments of 1984 (HSWA) made
many changes in Resource Conservation and Recovery Act (RCRA)
sections covering regulations of hazardous waste. The minimum
technology requirements of HSWA require EPA to revise regulations
for liners and leak detection systems at hazardous waste management
v
units. The Agency's minimum technology requirements for landfills,
surface impoundments, and waste piles require a double liner
system. The HSWA require an "approved leak detection system" to be
utilized at these new units. . The basis for the leak detection
system is the leachate collection and removal system (LCRS) between
the top and bottom liners as required in the regulations. The
ultimate goal of the liner and leak detection system is to prevent
the release of hazardous constituents from the unit.
The objective of the analyses described here is to simulate
the leakage from the top liner of the double liner system and the
movement of water through the underlying drainage layer to a drain.
The results of computer simulations are for use in the development
of action leakage rates (ALR). The action leakage rate is a
leakage rate that requires implementation of a response action to
prevent hazardous constituent migration out of the unit.
MODELING APPROACH
A three-dimensional finite element model, VAM3D, developed to
simulate water flow and solute transport in variably saturated
porous media was used to simulate flow from a punctured synthetic
upper liner to a drainage layer. The model was used to perform the
-------
three-dimensional simulations for the point source leak for both
landfill and surface impoundment cases. A series of simulations
were performed to investigate the effects of various input
parameters on the hydraulic head distribution and drain discharge
rates.
MODEL INPUT DATA
The following are the main - input data for performing the
simulation using a finite-element three-dimensional model:
. Thickness, length, width and slope of .the drainage layer;
location and area of the leak; hydraulic conductivity of
the drainage layer; location of the discharge drain; and
hydraulic heads at the leak and drain locations.
MODEL OUTPUT
The model calculates the distribution of hydraulic head in the
drainage layer and the flow rates through the leak in the liner and
the discharge to the drain.
MODELING ASSUMPTIONS
A model represents an idealization of a natural system.
Certain assumptions are necessary in making these representations.
The following are main assumptions underlying the analyses reported
here:
o Uniform properties throughout the drainage layer;
o Leak occurs through punctured hole of very small area
compared to the area of the landfill;
o Thickness of the drainage layer is uniformly one foot;
o The bottom, sides and upstream boundaries of modeled
region are impermeable;
o Steady state flow conditions prevail, and the flow in the
unsaturated zone is negligible;
-------
o Flow in the saturated zone occurs approximately parallel
to the slope of the bottom of the drainage layer;
o The water level in the drain is maintained at a constant
level near the bottom of the drainage layer;
o For the landfill case, the hydraulic head at the leak is
maintained at the top of the drainage layer for most of
the cases studied; and
o For the surface impoundment case, the hydraulic head at
the leak is maintained at the impounded water level.
SIMULATION SCENARIOS
A number of scenarios for representing the various waste
management units were selected for simulation:
1. Landfill and surface impoundment scenarios with a
leak in the top liner; head in the drainage layer
was kept at the top liner and other parameter were
varied to determine effects" on flow rates and head
distribution. The water thickness above the top
liner was more than zero for surface impoundments:
a. Distance of leak point from the drain;
b. Slope of the liner system;
c. Size of the leak; and
d. Hydraulic conductivity of the drainage layer.
2. Geonet scenario with similar parameter variations as for
the above scenarios.
SIMULATION RESULTS
A summary of results is presented here in a series of Figures
and Tables. The results are discussed in the Preamble to the
liner/leak detection rule for the development of ARL.
-------
LIST OF TABLES
Table Page
1. Flow Rates from Drainage Layer for Various Landfill
Scenarios (gal/day) 5
2.v Effects of b'ner Leak Size on Drain Discharge 6
3. Drainage Layer Flow Rates for Surface Impoundments.... 7
4. Drainage Layer. Flow Rates for Landfills and Surface
Impoundments 8
5. Flow Rates Through a Geonet 9
-------
Table 1. Flow Rates From Drainage Layer for Various
Landfill Scenarios (gal/day)
UJ
T
Leak 20 feet from drain
Case
(A)
(B)
(C)
(D)
Slope
0%
1%
2%
3%
Hydraulic Conductivity (K) in cm/sec
1
7394.2
8164.9
8485.8
8805.4
0.1
739.42
816.49
848.58
880.54
0.01
73.94
81.65
84.86
88.05
Landfill Area = 1 acre
Length Paralel to Flow = 100ft.
Leak Area = 1 sq. ft.
Thickness of Drainage Layer = 1 ft.
Head Above Top Liner = 0
Leak 80 feet from drain
Case/
Slope
(E) 0%
(F) 1%
(G) 2%
(H) 3%
Hydraulic Conductivity (cm/sec)
1 0.1 0.01
6075.7
8048.5
8564.0
8829.9
607.57
804.85
856.40
882.99
60.78
80.48
85.64
88.30
-------
Table 2.
Effects of Liner Leak Size
on Drain Discharge
Leak
Area
(sq. ft.)
0.123
0.175
0.25
1.0
3.0
4.0
12.5
20
25
30
Flow Rate
(GPD)
479
532
564
805
- 1337
1480
1700
1840
1950
1955
Area of Landfill = 1 acre
Hydraulic Conductivity = 0.1 cm/sec
Slope = 1 percent
Distance of leak from drain = 80 feet
Head above top liner = 0
-------
Table 3
Drain Flow Rates for Surface Impoundments
Hydraulic Head
above Top
Liner (ft.)
0
1
2
0
1
2
Slope
(%)
1
1
1 '
2
2
2
Flow Rate (gpd)
Hydraulic Conductivity (cm/sec.)
t 0.1 0.01
5640
15,400
27,190
5740
16,050
28,750
564
1540
2719
574
1605
2875
56
154
272
57
161
288
Area of waste management unit = 1 acre
Thickness of drainage layer = 1 ft
Distance from the leak = 80 ft
Leak area = 0.25 sq ft
f -32
-------
Table 4.
Drainage Layer Flow Rates For
Landfills and Surface Impoundments
Hydraulic Head
above Top
Liner (fti
1 '
1
0.5
0.5
Slope
(%\
1
2
1
2
Flow Rate tod)
Hydraulic Conductivity (cm/sec)
1 0.1 0.01
20,400
21,900
13,900
14,300
2040
2190
1390
1430
204
219
139
143
Thickness of Drainage Layer - 1 ft
Total area of waste management unit = 1 acre
Length of unit along flow direction - 200 ft
Distance of Leak from drain * 180 ft
Leak size « 1 sqft
-------
Table 5. Flow Rates Through a Geonet
Thickness of
Geonet
(mm)
5
5
5
5
5
5
5
5
10
10
10
10
Hydraulic Head
above Geonet
(ft.)
1
1
1
1
2
2
2
2
1
1
1
1
Slope
(%)
0
1
2
3
0
1
2
3
0
1
2
3
Flow Rate (gpd)
Hydraulic Conductivity (cm/sec.)
1 0.1 0.01
81
524
751
977
160
813
1040
1264
164
1625
2090
2552
8.1
52.4
75.1
97.7
16.0
81.3
104.0
126.4
16.4
162.5
209.0
255.2
0.8
5.2
7.5
9.8
1.6
8.1
10.4
12.6
1.6
16.3
20.9
25.5
Leak Area = 0.25 sq. ft.
41-35"'
-------
LIST OF FIGURES
Figure
1. Model Scenarios for Landfills and Surface Impoundments 11
2-a Water-Table Profiles in Drainage Layer Due to a Leak in
Top Liner (Cases A and E of Table 1 - No Slope) 12
2-b Water-Table Profiles in Drainage Layer Due to a Leak in
Top Liner (Cases F and B of Table 1 - 1%Slope) 13
2-c Water-Table Profiles in Drainage Layer Due to a Leak in
Top Liner (Cases C and G of Table 1 - 2% Slope) 14
2-d v Water-Table Profiles in Drainage Layer Due to a Leak in
Top Liner (Cases D and H of Table 1 -3% Slope) 15
3-a Water-Table Contours for Case F (Table 1) 16
3-b Water-Table Cdntours for Large Leak Area 17
4. Landfill Drainage Layer Flux Vs Drainage Layer Slope for
Different Leak Positions 18
5. Drain Discharge for Different Leak Sizes 19
6. Hydraulic Head Distribution in the Drainage Layer Due to
a Hydraulic Head of 2ft. Above Top Liner Leak Point 20
7. ' Comparison of Water-Table Profiles for Various Surface
Impoundment Scenarios 21
8-a Water Table Profiles in the Drainage Layer Due to a Leak in
Top Liner Under a Surface Impoundment (Case: 1%Slope) 22
8-b Water Table Profiles in the Drainage Layer Due to a Leak in
Top Liner Under a Surface Impoundment (Case: 2% Slope) 23
9 Thin Synthetic Drain Layer (GEONET) Scenario 24
10. Drain Flux Vs Slope for the Thin Synthetic Layer 25
-------
Rgure 1.
Model Scenarios for
Landfills or Surface Impoundments
Plan View
r
/
B'
Uok
Li-4*
\\V\\\N\ \ \N\ N \\\\ \\\\ \\\
Drain
Cross—section
Drain
dlschorge - Q
-------
Figure 2-a. Water-Table Profiles in Drainage Layer Due to a Leak in
Top Uner (Cases A and E of Table 1 - No Slope)
Case A : No Slope
-1
Distance in feet
Case E : No Slope
100
100
Distance in feet
-30-
-------
Figure 2-b. Water-Table Profiles in Drainage Layer Due to a Leak in
Top Liner (Cases F and B of Table 1 -1% Slope)
Case T i 1% Slope
-1
50
Distance, in feet
Case B : 1% Slope
50
Distance in feet
100
100
Note : Profiles do not change with changes in K,
Q changes proportionately.
-------
J-
Figure 2-c. Water-Table Profiles in Drainage Layer Due to a Leak in
Top Uner (Cases C and G of Table 1-2% Slope)
0 drain
Case C : 2% Slope
50
Distance in feet
100
-2--
Case G : 2% Slope
50
Distance in feet
100
-------
Figure 2-d. Water-Table Profiles in Drainage Layer Due to a Leak in
Top Liner (Cases D and H of Table 1-3% Slope)
Case D : 3% Slope
Q drain
100
Distance in f««t
Case H : 3% Slope
-3'
50
Distance in feet
100
-------
200
v(tt.) ;
150>
X
100
SO
\\
-
m ,
M
M
1*
•»
"/i
1
rrr
i
t
»
^
*-— *
^
1
> c
9 i
V
»
i:
i
> 5
4 4
! •
^
1
If
t *
1
j
.1.
1
> f
• c
11
1 1
* <
ft v
^
1
:i
l'
y
»
«
>
.
Figure 3-
i
^~ Drai
Line of s
• w »
Figure 3-a. Water Table Contours for Case F (Table 1)
Leak area = 1 sq. ft.
Slope m 1%
Distance of leak from drain
Contour interval« 0.1 ft.
= 80 ft.
Leak
50
too
-------
VWSXV VWW V
Y(ft.)
trr
H> ci * m o» >4 n
Figure 3-b. Water Table Contours for Large Leak Area
Drain
Line of symmetry
Leak
100
Leak area = 3 sq. ft.
Slope =1%
Distance of leak from drain
Contour interval « 0.1 ft
= 80 ft.
-------
Figure 4. Landfill Drainage Layer Flux Vs Drainage Layer Slope for
Different Leak Positions
•fc
90
Q
IX
O
or
80 J
70-
60-
50
0
K = 0.01 cm/s
leak 20 ft from drain
leak 80 ft from drain
2 4
Slope (in Percent)
-------
Rgure 5
Drain Discharge for Different Leak Sizes
2000
1600
1200
§.
O
ff
O
800
400
• •••••••••••••••••••••••••••••I
Hydrauflc ConductMty » 0.1 cm/i
Landfill ATM - 1 Acre
Head on Top liner - 0 ft
~.*
i i • • i i
10 15 20
Leak Area (sq ft)
30
-------
Figure 6. Hydraulic Head Distribution in the Drainage Layer Due to
a Hydraulic Head of 2 ft. Above Top Liner Leak Point
-1-
for K
Q(2')
Q(O')
0
0.1 cm/s
3552 GPD
80 GPD
100
Distance in feet
Surface Impoundment = 2 feet
Head at Leak = 0 feet
-------
-4
I
Figure 7. Comparison of Water-Table Profiles for Various Surface
Impoundment Scenarios
-1-
-2
Distance in feet
100
Surface Impoundment, Head = 2 ft. Case F [Q = 3552 GPD]
Surface Impoundment. Head = 2 ft on Length = 200 ft [Q = 3390 GPD]
Length of Domain = 200 ft (Head at Leak = 0) [Q = 707 GPD]
Base Case (P) (Head at Leak = 0) [Q = 805 GPD]
-------
Figure 8-a. Water Table Profiles in the Drainage Layer Due to a Leak ii
Top Uner Under a Surface Impoundment (Case: 1%Slope
Water Table Profile : 1« Slope
1
o
-1-
-2-
-3
Surfoct impoundment, H«od • 0.5 ft
Q - 1390 GPD
T—i—i—i—|—i—i—i—i—I—i—i—i—r
, 50 100
Distance in feet
Water Table Profile : Iff Slope
o
-2-
Surfoct Impoundment, Head • 1 ft
Q - 2040 GPD
-3-
-i—,
I i i i—i—f—i—r—i—T"
50 100 1
Distance in feet
-------
Figure 8-b. Water Table Profiles in the Drainage Uyer Due to a Leak in
Top Liner Under a Surface Impoundment (Case: 2% Slope)
s
Water Table Profile : 2* Slope
Surfoct Impoundment, Htad • 0.5 ft
Q • 1430 6PO
-4
1 I I T|)lll|lll*|ll*
0 50 100 150
Distance in feet
Water Table Profile : 2* Slope
^•M
(•ok
I
Surfact Impoundment, Htod • 1 ft
Q - 2180 GPD
-4
50 100 150
Distance In feet
-------
Figure 9.
Thin Synthetic Dram Layer (GEONET) Scenario
Uak
GEONET - Thin tynthttlc
drain layer
-------
Rgure 10
Drain Flux vs Slope for the Scenario of Thin Synthetic Layer
3000
2500
2000
1000
500
*7*
•
i^»* •••••••»•••
l
• •••••••••••••••••••
« ••••**••••
•••••• ••••*••••*••••*••••• •*•••
l
....... Head • 1 ft, Thickness - 5mm
- - Hsad - 2 ft, Thtcknsss - 5mm
— Head - 2 ft, Thtcknsss • 10mm
K - 1 cm/ssc., Leak Area - 0.5 sq
.••••••••^••••••••••••••••••••««**«*«**«*«*««*«****«»***«««»*****»#»*»**»»»»
• ••••••••
1
•••••••»••••»••<
••*••••••••••••••£••••••••••••••••••<
•••* »
..••
•••••*••••••••••••••
•••••••••••••••••••i
•••••••••••••••••••I
^--^J
lill
0.5
1.5
Slope (%)
-------
APPENDIX C
BACKGROUND INFORMATION ON THE
3-D VARIABLY SATURATED FLOW ANALYSIS MODEL
See also:
VAM2D—Variablv Saturated Analysis Model in Two Dimensions.
Version 5.2. Documentastion and User's Guide. NUREG/CR-5352,
Rev 1, HydroGeoLogic, Inc. for U.S. Nuclear Regulatory
Commission, October 1991.
Validation and Testing of the VAM2D Computer Code. NUREG/CR-
5795, HydroGeoLogic, Inc. for U.S. Nuclear Regulatory
Commission, October 1991.
-------
VAM3D-CG - Variably Saturated
Analysis Model in Three Dimensions
Version 2.3
HydroGeoLogic, Inc.
1165 Hemdon Parkway
Suite 900
Hemdon, VA 22070
Prepared for
Westinghouse Savannah River Co.
Atomic Energy Division
Savannah River Site
Aiken, SC 29808-0001
Westinghouse Project Manager : Ralph Nichols
-------
ABSTRACT
\
This report documents a three-dimensional finite element model, VAM3D-CG,
developed for simulating saturated-unsaturated groundwater flow and solute transport with
variable water table position and highly non-linear soil moisture conditions. The flow
equation is approximated using the Galerkin finite element method. VAM3D-CG has the
capacity to accommodate spatially variable hydraulic properties such as hydraulic
conductivity, storativity, and effective porosity, with high degrees of variability.
Nonlinearities due to unsaturated soil properties, atmospheric boundary conditions
V
(e.g., infiltration, evaporation, and seepage faces), and water uptake by plant roots are
treated using the Picard iteration technique or linearized using the Newton-Raphson scheme.
The transport equation may be>approximated using an upstream-weighted finite element
method to alleviate the problem of numerical oscillations. An orthogonal curvilinear
coordinate system may be used to discretize the domain, and elements can be designed along
subsurface formations.
Transport mechanisms considered include: advection, hydro-dynamic dispersion,
adsorption, and first-order decay. Complex boundary conditions, such as no-flow, constant
»
head, constant flux, constant gradient, and time-dependent head or flux, are easily
incorporated into the model. VAM3D-CG employs a Preconditioned Conjugate Gradient
(PCG) matrix solution scheme which allows several thousand nodal unknowns to be solved
extremely cost-effectively in transient or steady-state mode. The code can easily be adapted
for one-, two-, or three-dimensional applications, including axisvmmetric configurations.
-------
Several test problems are presented to verify the code and demonstrate its utility. Tnese
problems range from simple one-dimensional to complex three-dimensional problems.
-------
INTRODUCTION
BACKGROUND AND PURPOSE OF THE CODE
VAM3DCG is a three-dimensional, finite element code developed to simulate
moisture movement and solute transport in variably saturated porous media. The code is
capable of simulating a wide range of conditions commonly encountered in the field.
Simulations can be performed efficiently for fully three-dimensional, two-dimensional or
axisymmetric problems. Both flow and transport simulations can be handled concurrently or
sequentially. Material heterogeneities and anisotropy are handled by taking advantage of the
finite element approach. Efficient matrix computational and solution schemes are employed
in conjunction with simple rectangular prism elements to analyze problems involving highly
nonlinear soil moisture characteristics. Many types of boundary conditions can be
accommodated: 1) water table conditions, 2) atmospheric conditions associated with seepage
faces, evaporation and nonponding infiltration, 3) water uptake by plant roots, 4) vertical
recharge of the water table, and 5) pumping and injection wells.
The model formulation used in VAM3DCG is a descendant of the formulation used in
the FLAMINGO and VAM3D code presented by Huyakom, et al. (1986, 1987).
HydroGeoLogic, Inc. has recently enhanced certain portions of the published algorithms and
their coding to achieve greater flexibility, wider capability, and more robust numerical
performances when dealing with some difficult cases. Where possible, the new VAM3DCG
code has been rigorously checked against available analytical or semi-analytical solutions and
similar numerical codes including UNSAT2, FEMWATER/FEMWASTE, SATURN,
-------
FLAMINGO and VAM3D. A variety of field simulation problems described in the works of
Huyakorn et al. (1984,1985,1987), Enfield et al. (1983), and Carsel et al. (1985) have been
used to validate VAM3DCG and demonstrate is utility. Additional simulation problems are
described in this report.
OVERVIEW OF CODE CAPABILITIES AND SALIENT FEATURES
Multidimensional modeling of water flow and waste migration in variably saturated
subsurface systems can be a formidable task unless one is equipped with a proper code that
accommodates various field conditions. Recognizing this point, VAM3DCG was developed
to have not only essential modeling capabilities but also salient features mat facilitate
practical use. An overview of these aspects of the code is presented below.
1. VAM3DCG can perform transient analyses or single step steady-state analyses of
both variably saturated water flow and solute transport problems. If the flow and transport
problems are associated, a dual simulation can be made by solving the problems concurrently
or sequentially in a single computer run.
2. The finite element formulation and nonlinear solution procedures in VAM3DCG
are based on the state-of-the-art technology designed to accommodate a wide range of field
conditions including highly nonlinear moisture characteristics, material heterogeneity and
anisotrppy, and rapidly fluctuating transient flow boundary conditions.
3. VAM3DCG uses highly efficient matrix computational and matrix solution
techniques. The code is directly interfaced with newly developed ORTHOMIN and
Preconditioned Conjugate Gradient matrix solvers designed to handle problems with large
-------
number of nodal unknowns (on the order of several thousand or more) efficiently. This
feature makes the code attractive to use on a minicomputer or a personal computer PC 386.
4. An orthogonal curvilinear mesh can be used with this version of VAM3D-CG,
which makes the code attractive for undulating layered systems, and is better capable of
handling irregular boundaries, geometry, and material properties.
5. The flow simulator of VAM3DCG can handle various boundary conditions and
physical processes including infiltration, evaporation, plant root uptake, well pumping and
recharging, and varying water table conditions. Temporal variations of head and flux
boundary conditions can be handled conveniently using either continuous piecewise linear
*
representations or discontinuous (stepped) representations. The VAM3DCG code may also
be used as a modeling tool to supplement field investigation or other research study of
complex flow and transport behavior in variably saturated media.
6. The transport simulator of VAM3DCG is designed to handle both conservative
and nonconservative solutes. Its formulation is designed to have an upstream weighting
capability as an option to circumvent numerical oscillations. Both pulse and step releases of
contaminants from each source can be simulated.
APPLICABILITY OF THE CODE
The VAM3DCG code has many practical applications. Typical examples include the
following:
• Investigation of moisture movement and evapotranspiration in the
unsaturated zone including plant water uptake in the root zone.
-52-
-------
• Watershed studies - used to predict the response of x unconfined
watersheds to different schemes of drainage or to infiltration and
evaporation. The code computes spatial and temporal variations in
the pressure head, water saturation, and flow rate across specified
flow boundaries.
• Assessment of well performance and pumping test analy- sis - used to
analyze flow in the vicinity of pumped wells, to predict well
performance, and to prepare type curves for evaluation of pumping
test data.
• Mine dewatering investigations - used to predict the change in
elevations of water table or phreatic surface in response to mine
dewatering operations. These pre- dictions can be obtained by
performing area! or cross- sectional analyses of unconfined flow
problems.
• Contaminant migration assessment - used to predict leakage rates
and flow fields in unconfined aquifers underlying sewage ponds,
surface impoundments, and landfills. VAM3DCG can simulate
contaminant transport in variably or fully saturated porous media.
CODE USER REQUIREMENTS
In order to apply the VAM3DCG code effectively, the user
will need:
• a thorough understanding of hydrogeological principles
• a basic understanding of finite element techniques
« an awareness of the code's capabilities and limitations
• familiarity with the editor, operating system, and file handling concepts of the
computer system used.
It is also recommended that the user run some of the test problems provided to gain
confidence and understanding in using the code.
-------
COMPUTER EQUIPMENT REQUIREMENTS
VAM3DCG is written in ANSI Standard FORTRAN 77 and can be compiled on any
standard micro, mini, or mainframe system. The source code was developed and tested on
PRIME minicomputers and on PC386 micros using the FORTRAN 77 compiler developed by
the University of Salford, United Kingdom. With minor conversion (e.g.f changing OPEN
FILE statements), the source code can be made to compile and run on any machine equipped
with at least 2 megabytes of core memory, and a FORTRAN 77 compiler.
-------
EXAMPLE VERIFICATION AND APPLICATION PROBLEMS
Three sets of test problems were used for verification of numerical schemes and
demonstration of major capabilities of the VAM3DCG code. Specific purposes of the
example problems presented in this chapter are described below.
• Simulation of water flow under variably saturated (or saturated-
unsaturated) conditions
• Simulation of single component transport
• Coupled simulation of flow and transport problems with various
types of boundary conditions
• Verification of the VAM3DCG code against analytical solutions
and other finite element variably saturated flow and transport
codes (UNSAT2, FEMWATER and FEMWASTE)
• Demonstration of computational efficiency of the Preconditioned
Conjugate Gradient and ORTHOMIN algorithms implemented
into the VAM3DCG code
• Application of VAM3DCG to sample field problems.
The first problem set comprises four transient and steady-state flow problems with
different features of boundary conditions, dimensionality, and varying degree of nonlineariry.
These problems are as follows:
• Transient one-dimensional horizontal flow in a soil slab
• Transient vertical infiltration in a soil column
• Transient two-dimensional flow in a rectangular soil slab
• Steady three-dimensional flow in an unconfined aquifer with a pumping well.
-------
The second problem set comprises four transient transport problems. Three of these
problems are associated with three of the seven flow problems just mentioned. The transport
problems considered are listed as follows:
• One-dimensional horizontal transport in a soil slab
• Three-dimensional transport in uniform groundwater flow
• Two-dimensional transport in a rectangular soil slab
• Three-dimensional transport in an unconfined aquifer with a pumping welL
The third problem set comprises two associated flow and transport problems
concerning the simulation of moisture movement and contaminant migration in the
unsaturated zone surrounding a saltstone monolith in the z-area at the Savannah River Site,
South Carolina.
-62.-
-------
PROBLEM DEFINITION AND SIMULATION PROCEDURE
TYPES OF PROBLEMS
The VAM3DCG code can be used in several types of investigations of water flow and
moisture movement in •subsurface systems. For demonstrative purposes, four typical
examples are described. The first example (Figure 5.1) has application to the conceptual
design and risk assessment for a low-level radioactive waste disposal site. It involves
variably saturated flow around gravel and wick layers surrounding a low-level radioactive
waste container placed in the unsaturated zone above a water table (Frind et al., 1977). For
this study, VAM3DCG can be used to predict the flow pattern resulting from vertical
recharge at the soil surface. The velocity field determined from flow simulations can be
v
used as input to subsequent contaminant transport simulations. For Jhe investigation, or risk
analysis of the potential for migration of radionuclides, VAM3DCG can be used to perform
single-component transport simulations.
The second example (Figure 5.2) applies to drainage or mine dewatering problems
involving analyses of seepage into a drain or mine pit. For this example, VAM3DCG can be
used to'perform saturated-unsaturated flow simulations taking into account groundwater
recharge and drainage boundary conditions.
The third example applies to a landfill above an unconfined groundwater system
intercepting a river (Figure 5.3). To evaluate the environmental impact of a land disposal
unit (landfill or surface impoundment), it is essential to predict water flow and contaminant
-L3-
-------
i- STABLE SOIL SURFACE
J.I 1 I i.
.i^r^o RAVEL
*••.*•: i * %*m m
*„••»% LA Yen
f S f r
, CONCRETE '
^ CONTAINER '
WATER TABLE
Figure 5.1. Basic geologic environment for a low-level waste container (Frind et ah,
1977).
-------
LAND SURFACE
PIT MINE
f t / / / /
IMPERMEABLE
'/ "*
// /
AQUIFER
LAND SURFACE
— -
nun
MWNG EXCAVATION
[ t t t t t 1
—
(b)
Rgure 5.2. Groundwater seepage due to mine dewatering or underground drainage
operations.
-------
WATER TABLE
RIVER;
FLOW-
Figure 5.3. Groundwater contamination caused by a landfill.
-------
migration in both unsaturated and saturated zones between the landfill and the river.
VAM3DCG can be used to perform both the flow and transport simulations.
The fourth example concerns soil and groundwater contamination problems due to
application of pesticides. In the situation depicted in Figure 5.4), VAM3DCO may be used
to provide coupled transient simulations of moisture movement, groundwater flow and
pesticide migration through the root zone, the vadose zone and the saturated zone of
unconfined aquifer system. If chemical transformation and chained reactions of pesticides
are important, the code may also be used to study these effects on the fate and transport
-------
PESTICIDE APPLICATION
/"'RECEPTOR WELL
Figure 5.4. Soil and groundwater contamination caused by application of pesticides.
-6*-
-------
Frind, E.O., R.W. Gillham, and J,F. Pickens, 1977. Application of unsatunted flow
properties in the design of geologic environments for radioactive waste storage
facilities, Finite Elements in Water Resources, edited by W.G. Gray, G.F.
Finder, and C.A. Brebbia, Pentech Press, Plymouth, England, pp. 3.133-
3.163.
Huyakorn, P.S., P.F. Andersen, J.W. Mercer, and H.O. White, Jr., 1987. Saltwater
intrusion in aqufiers: development and testing of a three-dimensional finite
element model, Water Resources Research, 23, 2, 293-312.
Huyakorn, P.S., E.P. Springer, V. Guvanasen, and T.D. Wadsworth, 1986. A three-
dimensional finite element model for simulating water flow in variably
saturated porous media, Water Resources Research, 22,12,1790-1808.
Huyakorn, P.S., J.W. Mercer, and D.S. Ward, 1985. Finite element matrix and
mass balance computational schemes for transport in variably saturated porous
media, Water Resour. Res., v. 21, no. 3, pp. 346-358.
Huyakorn, P.S., S.D. Thomas, and B.M. Thompson, 1984. Techniques for making
finite elements competitive in modeling flow in variably saturated porous
media, Water Resour. Res., v. 20, no. 8, pp. 1099-1115.
------- |