| 

Landfill-gas
system design is a series of iterative steps often more
of an art than a science.
By Daniel P. Duffy
The very
nature of disposed solid waste, with its varying characteristics
and inconsistencies, makes it impossible to precisely
predict the amount and rate of landfill gas (LFG) production.
Therefore, the design of a LFG system should have a
built-in flexibility to allow for field changes and
layout modifications.
Predicting
Production
Waste arrives at a landfill with a typical
density of 33 pounds per cubic foot or 900 pounds per
cubic yard. This waste is usually compacted to an in-place
density twice that of its delivered density, about 67
pounds per cubic foot or 1,800 pounds per cubic yard.
The in-place waste is assumed to have typical “soil
characteristics” for municipal solid waste. These
characteristics are the same characteristics used to
predict leachate formation and precipitation percolation
rates as well as LFG production. Waste is assumed to
have a typical field capacity (the amount of water in
a “soil” which remains after extensive gravity
drainage) of 30% by volume and a total porosity (volume
of voids/total volume) of 67%. In certain areas of the
landfill, liquid content may reach total porosity, leading
to a saturated condition and contributing to leachate
outbreaks.
In a completely
saturated landfill strata, little if any LFG is produced.
Therefore the moisture content of the waste can greatly
affect the projected gas volumes.
These characteristics
determine the LFG production potential of the in-place
waste. Since these can vary significantly from landfill
to landfill and even within the same landfill, these
assumed characteristics can only be used as a rough
average for planning purposes. All subsequent computations
and predictions derived from these assumed characteristics
represent a “best guess” only. The LFG management
system design will almost certainly have to be modified
to accommodate actual field conditions.
The single
most important factor in predicting LFG production rates
is the rate of decomposition of the organic components
of the waste.
The available
decomposable mass of refuse for any year is given by
the following equation:
V1 = V0 + D1
where,
V1 = available decomposable mass for the year (cubic
yards converted to lbs)
V0 = remaining decomposable mass from the previous year
(cubic yards converted to lbs)
D1 = refuse disposed for the year (cubic yards converted
to lbs)
The amount of LFG produced per year is estimated by
the following:
P1 = V1 * (0.10 cubic feet per lbs)
where,
P1 = annual LFG production rate (cubic feet per year)
V1 = available decomposable mass for the year (cubic
yards converted to lbs)
The value
of 0.1 cubic feet of LFG per pound of in place waste
is empirical and based on field observations.
But since
the waste characteristics even within a landfill can
vary widely, this remains a rough average for planning
purposes. Further dividing P1 by 525,600 minutes per
year gives the average LFG production rate per minute.
The annual
amount of waste mass decomposed by the LFG production
process is given by the following:
dV1 = P1 / 4.49
where,
dV1 = the amount of waste converted to LFG (lbs)
P1 = annual LFG production rate (cubic feet per year)
The factor
of 4.49 converts the cubic feet of LFG into lbs of decomposed
waste. Its value is based on laboratory studies and
is an average value subject to change according to local
waste characteristics. The remaining decomposable waste
carried over to the next year’s LFG product estimate
is given by the following:
V2 = V1 - dV1
where,
V2 = decomposable waste mass available for next year
(lbs)
V1 = available decomposable mass for the year (lbs)
dV1 = the amount of waste converted to LFG (lbs)
A hypothetical
landfill with a total disposal capacity of 1,850,000
cubic yards (equivalent to about 1,150 acre-feet) might
utilize an average of 92,500 cubic yards of airspace
per year over a 20-year operational lifetime.
With an
average in-place density of 1,800 lbs per cubic yard,
the landfill would receive an average of 82,750 tons
per year (or 230 tons per day, typical for a moderate-size
landfill). The maximum in-place waste volume is achieved
at the end of year 20, when disposal operations cease.
Initial
LFG production from the first year of waste disposal
would be a relatively small 32 cubic feet per minute.
LFG production
also peaks at the end of year 20 with a production rate
of 516 cubic feet per minute. It then declines over
the course of the site’s 30-year, post-closure
care period to a rate of 216 cubic feet per minute.
The annual
average rate of LFG production is shown in Table 1.
| Table 1 - Estimated LFG Production Rates |
| Yr |
Disposal (tons) |
In Place Waste (tons) |
Decomposed (tons) |
LFG (cfm) |
| 1 |
83250 |
83250 |
1854 |
32 |
| 5 |
83250 |
398117 |
8867 |
151 |
| 10 |
83250 |
75831 |
16789 |
287 |
| 15 |
83250 |
1071660 |
23868 |
408 |
| 20 |
83250 |
1355638 |
30192 |
516 |
| 25 |
0 |
1211252 |
26977 |
461 |
| 30 |
0 |
1082244 |
24103 |
412 |
| 35 |
0 |
966977 |
21536 |
368 |
| 40 |
0 |
863987 |
19242 |
329 |
| 45 |
0 |
771966 |
17193 |
294 |
| 50 |
0 |
689746 |
15362 |
262 |
Final
Cap and Cover
The efficiency of the LFG extraction system
is primarily dependent on the permeability of the cap.
The amount of gas that can be extracted per well, and
the spacing between the wells, can be greatly increased
with an impermeable cap.
The greater
the spacing between the wells, the fewer wells that
are required to ensure complete coverage of the landfill.
This results in significant capital cost savings and
long-term savings in maintenance costs.
For our
hypothetical landfill we will assume the following final
cover system (from top to bottom):
- 6 inches
of topsoil (soil type SM, silty sands with a permeability
of 5.2 x 10-4 cm/sec)
- 24 inches
of soil cover (soil type SM, gravelly sands with a
permeability of 1.7 x 10-3 cm/sec)
- 0.24
inches thick geocomposite lateral drainage net (thick
factory bonded geotextile-geonet-geotextile)
- 40 mil
high-density polyethylene geomembrane (HDPE) on the
top. The HDPE is factory-bonded to the underlying
GCL. The edges of the HDPE geomembrane panels may
not be seamed during installation, and therefore it
is assumed the geomembrane’s liner leakage fraction
is greatly increased.
- 0.20-inches-thick
composite barrier layer. A geosynthetic clay liner
(GCL) consisting of a bentonite layer (permeability
of 5.0 x 10-9 cm/sec) sandwiched between a geotextile
layer on the bottom and
- 12 inches
of grading soil (soil type SW, well graded gravelly
sands with a permeability of 5.8 x 10-3 cm/sec)
Each hydraulic
conductivity results in an equivalent relative cover
factor (Ms). This factor and the thickness of the layer
are used to determine the final cap’s overall
impermeability to gas.
Table 2
summarizes the relationship between the cover material’s
vertical hydraulic conductivities and the relative cover
factor, Ms. Table 3 summarizes the configuration and
material characteristics of the final cover system.
| Table 2 - Relative Cover Premeability Factors |
| Material Vertical Hydraulic Conductivity (cm/sec) |
Relative Cover Factor, MS |
| 1.00 x 103, or more |
0.10 |
| 1.00 x 10-4 |
0.40 |
| 1.00 x 10-5 |
0.70 |
| 1.00 x 10-6 |
0.90 |
| 1.00 x 10-7 |
1.00 |
| 1.00 x 10-8 |
1.10 |
| 1.00 x 10-9 |
1.20 |
| Table 3 - Final Cover Configuration and Material Characteristics |
| Layer |
Thickness (fee) |
Vertical Hydraulic Conductivity (cm/sec) |
| Topsoil |
0.50 |
5.2 x 10-4 |
| Protective Cover Soil |
2.00 |
1.7 x 10-3 |
| Drainage Layer |
0.24 |
NA |
| 40 ml HDPE |
0.0025 |
2.0 x 10-13 |
| GCL |
0.20 |
5.0 x 10-9 |
| Grading Soil |
12.00 |
5.8 x 10-3 |
Extraction
Wells Depths and Zones of Control
Active landfill gas extraction wells are typically
located within the limits of waste and set at depths
of approximately 75% of the waste thickness at the well
locations. The deeper wells are located in the center,
with shallower wells around the waste perimeter. Since
the well’s zone of influence (ZOI) changes with
depth, and depth changes with location, several trial
iterations using different well locations and depths
may be required before the design can ensure complete
coverage of the limits of waste. Once the well length
has been determined, usually two thirds of its length
is kept solid while the bottom one third of its length
is slotted to allow inflow of the LFG.
Table 4
summarizes the relationship between the hypothetical
depths of waste, the total lengths of the LFG extraction
wells and the slotted and solid pipe segments of the
wells.
| Table 4 - Waste Depth and WEll Length Relationships |
| Item |
Formula |
Length (feet) |
| Depth of Waste at Well Location, Dw |
|
40 |
| LFG Well Length, Lw |
= Dw * 75% |
30 |
| Solid Pipe Length of the LFG Well, Sp |
= Lw * 67% |
20 |
| Slotted Pipe Length of theLFG Well, Sl |
= Lw * 33% |
10 |
Tables 5,
6, and 7 depict equivalent thickness of the geomembrane,
relative cover permeability, and LFG well Zone of Influence
(ZOI), respectively.
| Table 5 - Equivalent Thickness of the Geomembranes |
| Item |
Value |
| Vertical Hydraulic Conductivity of compacted clay layer (cm/sec), KC = |
0.000001 |
| Vertical Hydraulic Conductivity of the Geomembrane(cm/sec), KH = |
0.0000000000002 |
| Thickness of the 30 mil Geomembrane (feet) , Cd H = |
0.0025 |
| Liner Leakage Fraction (dimensionless), LLF = |
0.004 |
| Equivalent Geomembrane (feet), Cd E =( KC/KH)*(CdH*LLF)= |
5.00 |
| Table 6 - Relative Cover Permeability |
| Layer |
Cover Depth (feet), Cd |
Vertical Hydraulic Conductivity, K |
Permeability Factor, Ms |
Relative Cover, Ms'=Ms*Cd |
| 1 |
0.50 |
0.00052 |
0.40 |
0.20 |
| 5 |
2.00 |
0.0017 |
0.10 |
0.25 |
| 10 |
NA |
NA |
NA |
0.00 |
| 15 |
5.00 |
0.000000000002 |
1.20 |
6.00 |
| 20 |
0.20 |
0.000000005 |
1.20 |
0.24 |
| 25 |
1.0 |
0.0058 |
0.10 |
0.10 |
| 30 |
8.70 |
|
|
6.79 |
| Table 7 - LFG Well Z01 |
| Item |
Value |
| Refuse Permeability Factor, F S = |
4.00 |
| Solid Pipe Length of the LFG Well, Sp = |
20.00 |
| Relative Cover Permeability Factor, Ms = |
6.79 |
| Total Cover Depth, Cd = |
8.70 |
| LFG Well Zone of Influenc (feet),ZOI =Fs*(Sp+[Ms*Cd]) = |
316.29 |
LFG well
ZOIs are estimated by the following formula:
ZOI = Fs * (Sp+[Ms * Cd])
where,
ZOI = LFG extraction well zone of influence (feet)
Fs = Refuse permeability factor (dimensionless, range
from 3.5 to 6.5)
Sp = Length of the solid pipe section of the LFG well
(feet)
Ms = Relative cover permeability factor (dimensionless)
Cd = Cover depth (feet)
If the final
cover system consists of multiple layers of soil and
geosynthetics, the relative cover permeability factor
for the cover system as a whole is determined by the
following:
Ms = (Ms1 * Cd1) + (Ms2 * Cd2) +
… + (MsN * CdN)
The equivalent
thickness of a geomembrane portion of a composite cap
is determined in relation to the clay component of a
composite cap, and is calculated by the following:
CdE = (KC / KH) * (CdH * LLF)
where,
CdE = Equivalent thickness of the geomembrane (feet)
KC = Vertical hydraulic conductivity of the clay (cm/sec)
KH = Vertical hydraulic conductivity of the geomembrane
(cm/sec)
CdH = Thickness of the geomembrane (mils, converted
to feet)
LLF = Liner leakage fraction (dimensionless, from HELP
Model analyses)
(Note: for
the purposes of determining ZOI, the equivalent thickness
of the geomembrane assumes that the edges of the geomembrane
are not welded and therefore significantly increase
the geomembrane’s liner leakage fraction.)
The thickness
of a geocomposite (factory bonded geotextile-geonet-geotextile)
drainage blanket in the final cover system is not included
in the analysis. The equivalent thickness of the geomembrane
portion of the hypothetical cap is computed as follows.
For the
purposes of design, a conservative ZOI of 300.0 feet
will be used for each proposed LFG extraction well set
in waste having a thickness of 40 feet or less. Note
that if the final cover system did not use a geomembrane
cap in its cover layer, its relative cover permeability
factor would be reduced from 6.79 to 0.79, a reduction
of 88%.
This should
illustrate the importance of a geomembrane cap to an
efficient LFG extraction system. A geomembrane acts
to ensure the maintenance of pressure within a landfill
and to prevent air infiltration into the landfill as
a result of suction pressure from the wells.
Without
the geomembrane, the resulting ZOI per well would be
much smaller and require many more wells to provide
full coverage.
Now that
the wells have been properly spaced for maximum coverage,
we need to determine the reduction in their areas of
influence resulting form overlapping, adjacent ZOIs.
Each overlap area is determined by the following formula:
Sx = [R2 * COS-1 (d/R)]
– [d * (R2– d2)_]
where,
Sx = the overlap area (square feet)
R = LFG well ZOI (feet)
d = half the distance between the adjacent gas wells
(feet)
The flow reduction factor is determined by the following
formula:
Rf = (Sx1 + Sx2 + ... + Sxn) / (pi * R2)
Since each
LFG extraction well ZOI will be overlapped by up to
six adjacent ZOIs, each overlap area (Sx) cannot exceed
2.5% of the total well area of influence to ensure the
maximum allowable Rf of 15% is not exceeded. If a well
is located in a corner of a landfill or along its perimeter
(and does not have six adjacent wells) the overlap per
adjacent well can be proportionally higher, just so
long as the 15% total is not exceeded. In our hypothetical
example, well spacing of 530 feet for wells having ZOIs
of 300 feet will provide sufficient coverage without
excessive overlap.
LFG
Pipe Flows, Velocities and Head Losses
After the wells have been properly spaced,
the amount of gas extracted by each well can be determined.
This is
done by first estimating the volume of in-place waste
within the reduced ZOI of each well.
Each LFG
extraction well influences a cylindrical volume of in-place
waste. It is from this volume that the LFG is extracted.
The volume of this cylinder is determined by the following
formula:
V = pi *
(ZOI2 ) * Dw * (100% – Rf )
where,
V = Volume of influence (cubic feet)
ZOI = Zone of influence (feet)
Dw = Depth of waste (feet)
Rf = Flow reduction factor (percent)
The extraction rate from this volume of influence is
a function of the gas generation rate of the in-place
waste:
Qu = V * (dg / dt) * dw
where,
Qu = LFG extraction rate (cfm)
V = Volume of Influence (cubic feet)
dg/dt = LFG generation rate per lbs of in-place waste
(cubic feet/[lbs * minute] )
dw = in-place density of waste (lbs/cubic feet)
For our
hypothetical example, each LFG extraction well is estimated
to have an extraction flow rate of approximately 90
cfm. Dividing the estimated peak LFG production rate
(previously calculated) by this value will provide the
minimum number of wells required.
An examination
of the design plan showing the well locations derived
from the required well spacing will determine if the
proposed design has the minimum number of required wells.
In order
to function properly, each LFG extraction wellhead needs
a minimum pressure head of 10 inches of water column.
Since the pressure is applied by a single, central blower,
the pressure-head losses in the lateral pipelines connecting
the wells to the blower need to be determined. These
lateral pipes are usually solid, high density polyethylene
(HDPE) with diameters varying from 6 inches to 24 inches.
There are
two ways of installing lateral and header pipelines,
either above or below the geosynthetic components of
the final cover system.
Installation
above the geosynthetics is a more difficult construction
effort, the components of the final cover all have to
be graded and trenched along the alignment of the pipelines
well before they are installed.
Furthermore,
the need to pre-align the pipelines severely restricts
operational flexibility and field modifications to the
system.
However,
it makes repair and maintenance of the pipelines much
easier. Installation below the geosynthetics has the
reverse characteristics, ease of construction and operational
flexibility, but with later difficulty (due to the need
to slice through the overlaying geosynthetics) with
repair and maintenance.
Of the two,
the below-the-geosynthetic installation is preferred
since all LFG estimates are best guesses only, and operational
flexibility allowing for field changes must be retained.
Pressure
drops for gaseous pipeline flows are approximated by
the Spitzglass equation for relatively low pressure
flows:
P = G *
L * ( Q/[59.167 * K])2
where,
P = Pressure drop (inches of water column)
G = Specific gravity of the LFG (0.98, dimensionless)
L = Length of pipe segment (feet)
Q = Volumetric flow rate (cfm)
K = Spitzglass Constant (dimensionless)
When determining
head losses along a header pipeline, each wellhead connected
to the main header pipeline by a lateral pipe is considered
a node at a point along the header. The header itself
is usually circular so that if blockage occurs at a
point along its length, LFG can still be extracted in
the opposite direction. Though this will not occur at
the same pressure head or with the same efficiencies,
it is better than completely blocked gas flows. That
can result in a disastrous and potentially dangerous
expansion of the geomembrane cap (blowing up like a
balloon) as the gas accumulates behind the blockage.
Though minor
head losses from pipe bends, tees, flanges, and other
fixtures is insignificant and usually ignored, there
is an additional pressure-head loss in the flare stack
used to flare off the extracted gas. This head loss
is typically 12 inches of water column. As a safety
factor, an additional 10% is added to the minimum required
head. So if a maximum pipeline head loss of 4 inches
occurs between the blower and the farthest extraction
well, a minimum 28 inches of water column will be required
for the system with the factor of safety. This is equal
to 145 psf or over 1 psi. The blower should be rated
for this pressure and the maximum gaseous flow rate.
The pipe
diameters should be checked to ensure that the gaseous
flow velocities do not exceed the maximum allowable.
The maximum allowable gas flow velocity in a direction
concurrent with condensate flows is 40 feet per second.
For counter concurrent flows, the maximum velocity is
20 feet per second.
Condensate
Collection and Control
Condensate flows are liquid flows that follow
the slope gradient of the pipeline. Gas extracted by
pressure follows the pressure gradient of the pipeline.
Often the two are in opposing directions.
Should the
pipe diameters need adjusting, the head losses in the
pipeline must be recalculated. Then the velocities are
checked again. This iterative process continues until
an acceptable design is achieved.
Condensate
is a particularly nasty and concentrated form of leachate
that condenses out of the landfill gas as it cools while
traveling along the pipeline. It is important to remove
this accumulated liquid, as it can cause pipeline blockage.
Condensate blockages are evidenced by a “sloshing”
sound and indicated by rapid fluctuations in gas flow
rate, temperature and increasing negative pressure.
Condensate
accumulates and must be removed from low elevation points
along the header pipeline.
While the
condensate itself has a typical pH range of 3.5 to 7.5,
it contains highly acidic compounds (at concentrations
as high as 4,000 ppm). The acidic compounds that are
captured by the condensate can give rise to relatively
high rates of corrosion of carbon-steel pipe and fittings.
Typical condensate also includes relatively high proportions
of chlorides, ammonia nitrogen and phenols. Some states
require the condensate to be collected and treated separately.
This would
require an additional system of pipelines to convey
the condensate from the low collection points in the
header pipeline to an onsite pretreatment facility.
Most states
however, allow the condensate to be recycled back into
the landfill. Compared with the amount of leachate collected
from the bottom of the landfill the amount of condensate
collected from the LFG system is relatively small. The
small quantities of condensate get diffused in the leachate
and have there constituents mitigated by percolating
through the waste.
Environmental
engineer Daniel P. Duffy, PE, lives and works in Cincinnati, OH
MSW
- March/April 2006 |