-------

-------
capital cost.  Fuel cells  are  an  emerging technology that may ultimately
improve the outlook for clean,  efficient, and economical energy use of landfill
gas.

      Early 1995 marked the  completion of a major step toward establishing
fuel cells as a promising  new  technology for utilizing landfill gas.  The first
demonstration using landfill gas  in a commercially available phosphoric  acid
fuel cell power plant was  successfully conducted at the Penrose Landfill in Sun
Valley, California.  This  was  the conclusion of a three-phase program which
began in 1990 when EPA awarded International Fuel Cells Corporation- (IFC) a
contract to demonstrate landfill  gas control with energy recovery.  The project
addressed two principal issues: (1) a landfill gas cleanup method to remove
contaminants from the gas  sufficiently for fuel cell operation, and (2) a
demonstration test of a commercial fuel cell power plant operating on landfill
gas.

      This paper will summarize some of the latest test results and the status
of the project.  Additional  information can be found in EPA reports [2], [3].

LANDFILL  GAS  FUEL  CELL

Overall System

      Figure 1 provides a  simplified description of the required components of
a fuel cell landfill gas-to-energy system.  The system consists of the landfill
gas wells and collection system,  a modular gas pretreatment unit (GPU), and a
fuel cell power plant modified for landfill gas operation.  The fuel cell power
plant is an adaptation from the mitural-gas-fueled PC25 fuel cell sold by ONSI
Corporation,  an IFC subsidiary.   Landfill gas collected at the site is
processed to remove contaminants  in the GPU.  The cleaned medium heat content
landfill gas is then used  to fuel the fuel cell power plant to produce
alterating current (AC)  for sale  to the electric utility.  The clean
cogeneration heat produced by the fuel cell may also be utilized if a
requirement for heat exists at the site.  If not utilized, it is rejected by
the air cooling module.

-------
                                                                      UUUty
                                                                      power
        Undiii
                   FIGURE  1.  FUEL CELL LANDFILL GAS ENERGY RECOVERY
Fuel Cell Power Plant
      A simplified- functional schematic of the 200 kW fuel cell power unit is
shown in Figure 2.  Major sections of the system, include the fuel processing
system, fuel cell stack,  and the thermal management system.  In the fuel
processing section treated landfill gas is converted to hydrogen (H2)  and
carbon dioxide (CO2) for introduction into fuel cell stack.  The fuel  treatment
process includes a low temperature fuel preprocessor to remove the residual
contaminants from the treated gas, a fuel reformer, and a low temperature shift
converter where the exhaust from the reformer is further processed to provide
additional H2 and C02.

      In fhe fuel cell stack, H2 from the process fuel stream is combined
electrochemically with oxygen from the air to produce direct current  (DC)
electricity and byproduct water.   The product water is recovered and used in
the reformer.  The heat generated in the cell stack is removed to an external
heat rejection system.  This energy can be either rejected .to the ambient air.

-------
or recovered for use by the customer.   The DC power produced in a fuel cell
stack is converted to AC power in a  power conditioning package not shown on the
process schematic.
                                   MANAGEMENT
                                   SYSTEM
       FIGURE 2. FUNCTIONAL SCHEMATIC FUEL CELL LANDFILL GAS  POWER UNIT

-------
      A  preliminary design of a fuel cell  power plant was established to
 identify the  design requirements which allow optimum operation on landfill  gas.
 Three issues  specific to landfill gas operation were identified which reflect  a
 departure from a- design optimized for operating on natural gas.  A primary •
 issue is to protect the fuel cell from sulfur and halide compounds in the
 landfill gas.  A second issue is to provide  mechanical components in the
 reactant gas  supply systems to accommodate the large flow rates that result
 from use of dilute methane fuel.  The third  issue is an increase in the heat
 rate of  the power plant by approximately 10% above that anticipated from
 operation on  natural gas.  This is a result  of the inefficiency of using the
 dilute methane fuel.  The efficiency results in an increase in heat recoverable
 from the power plant.  Because the effective fuel cost is relatively low, this
 decrease in power plant efficiency will not  have a significant impact on the
 overall  power plant economics.

      The landfill gas power plant design  provides a packaged, truck
 transportable, self-contained fuel cell power plant  with a continuous
 electrical rating of 200 kW.  It is designed for automatic, unattended
 operation, and can be remotely monitored.  It can power electrical loads either
 in parallel with the utility grid or isolated from the grid.

      In summary, a landfill gas fuel power  plant can be designed to provide
 200 kW of electric output without need for technology developments..  The design
would require selected components to increase reactant flow rates with a
minimum  pressure drop.  To implement the design would require non-recurring
expenses  for  system and component design,  verification testing of the new
components, and system testing to verify the power plant performance and
overall  system integration.   A thermodynamic analysis of the fuel cell power
plant optimized for operating on landfill  gas  was completed.  The resulting
performance of the landfill, gas power plant  is-compared to a power plant
operating on natural gas in Table 1.

-------
           TABLE 1. PERFORMANCE  COMPARISON  FOR NOMINAL  200  KW OUTPUT

Fuel
Electrical Efficiency (LHV) - % . .
Heat Rate (HHV) -kcal/kWhr
Available Heat - kcal/hr
Ambient Temperature for Fuel Water Recovery - °C
Startup Fuel
Natural Gas
Power Plant
Natural Gas
40
2,390
192,000
35
Natural Gas
Landfill Gas
Power Plant
Landfill Gas
36.4
2,620
208,000
35
Natural Gas
Landfill Gas Cleanup

      Landfill gases consist primarily of CO2', methane (CH4) ,  and nitrogen (N2) ,
plus trace amounts of hydrogen sulfide  (H2S),  organic sulfur,  organic halides,
and non-methane hydrocarbons.  N2 is not a true constituent of landfill gas,
but is drawn into the mix when vacuuming gas  into the collection  system.   The
concentration varies widely, with highest concentrations occurring  from the
landfill's perimeter wells.  The specific contaminants in the landfill  gas of
concern to the fuel cell are sulfur and halides.  Both of these components can
"poison" and, therefore, reduce the life of the power plant's fuel processor.
The fuel processor is the unit which converts CH4 in the landfill gas into H2
and CO2  in  an endothermic reaction over a catalyst bed.   The catalyst in this
bed can react with the halides and sulfides and lose its activity.  This
reaction,  when it occurs, is irreversible.

      The GPU, designed to remove fuel cell contaminants, is shown in Figure 3.
H2S is first  removed  by  adsorption on a packed .bed.   The material which
performs this function is a specially treated carbon, activated to catalyze  the
conversion of H2S  into elemental  sulfur which is deposited on  the bed.   The
reactions for the conversion of sulfur, the Claus reactions, are:

            H2S +  3/2 Oa  --> H2O + SO2

           , SO2 +  2H2S --> 2H2O + 3S
The bed is not regenerable on site, but must be removed to another site  if
regeneration is desired.

-------
UMM
1
L






F1U.QA8
HI*
MCMOUU.






^







eaoun
COHMNSI8


RAM



^i


I^H»




BHT1H
•CD
•"JBW?0"
CYCLE)

ORVCR
nHt«Sa
-------
regenerable dryer and hydrocarbon removal beds operate on a nominal 16-hour
cycle, with each set of beds  operating  in the adsorption mode for 8 hours and
in the regeneration mode  for  8  hours.

      The gas then passes through a particulate filter and is warmed indirectly
by an ambient air, finned-tube  heat exchanger to attain a temperature  above  O°C
before being fed to the fuel  cell unit.

GPU  TEST  RESULTS

      The GPU depicted in Figure 3 has  been successfully tested.  Detailed
results of the tests, along with the test plan, are contained in an EPA report
[3] .  The results of those tests will be summarized here.

      After completing 216 hours of continuous operation and a total of 616
hours since the first GPU startup, performance testing was conducted over a  3
day period at the beginning,  middle, and end of the regenerative cycle to
evaluate performance over the normal 8-hour cycles on each of the two
regenerative beds (dryer bed  and carbon bed).  At specific times in the
regeneration cycles,  Tedlar bag samples were collected from sampling manifolds
located at the GPU inlet and  outlet as  well as at the flare inlet and outlet.
These bag samples were analyzed off-site using gas chromatography/mass
spectrometry (GC/MS)  analysis for the volatile organic compounds (VOCs) and
gas chromatography/flame photometric detection (GC/FPD) analysis for sulfur
compounds.   Additionally,  sulfur compounds were measured at the GPU's inlet  and
outlet using on-line GC/FPD.  No on-line or on-site measurements were utilized
for VOCs because it was found that the  landfill gas matrix could bias the
results.  This was determined via^-an audit using certified cylinder gases.:.

      The results from one of the cycles are summarized in Table 2.   All the
measurements demonstrated that the GPU  was very effective in removing the
target sulfur compounds and VOCs.  For  sulfur compounds,  the GPU outlet
concentrations were either below detection limits (0.01 ppm for the on-line
method and 0.004 ppm for the  off-site analyses)  or in the parts per billion
concentration range.   Likewise halogenated and other VOCs were below 0.002 ppm
with the exception of methylene chloride, which was detected in trace levels of
less than 0.02 ppm.

      The fuel test data for  the GPU flare have been summarized elsewhere [3] .
It suffices to state that the flare destruction of VOCs and sulfur compounds
exceeded 99%.   Nitrogen oxides and carbon monoxide concentrations at the flare

-------
 outlet average 10.4 and 3.0  ppm,  respectively,  over  three  test  periods.
 Particulate matter averaged  0.03  mg/m3.


                     TABLE  2.  GPU  INLET/OUTLET  EMISSION TEST  DATA


Sampling Location

Reduced Sulfur Compounds (ppm)
Sample Type
hydrogen sulfide
carbonyl sulfide
methyl mercaptan
ethyl mercaptan
dimethyl sulfide
carbon disulfide
dimethyl disulfide
Total Reduced Sulfur - see note 1
Volatile Organic Halogens -
GC/US Analysis (ppm)
Sample Type
Compound
dichlorodifluoromethane
vinyl chloride
methylene chloride
cis- 1.2-dichloroethene
1, 1-dichloroethane
trichloroetnene
tetrachloroethene
chlorobenzene
Total Halogens (as halide) - see note 2
Volatile Organic Compounds -
GC/MS Analysis (ppm)
benzene
toluene
xylenes
ethyl benzene
styrene
acetone
2-butanone
ethyl acetate
ethyl butyrate
alpha-pinene
d-limonene
tetrahydrofuran
1800-1900
Hour 1
GPU
Inlet

bag
92.7
0.197
2.91
0.48
6.51
<0.07
<0.07
104

bag
0.83
1.1
6.6
4.3
1.9
1.3
2.7
0.91
46.7

1.1
28
14.9
6.1
0.6
<1.2
5.2
8.1
6.3
10.8
12.6
1.3
GPU
Outlet

bag
<0.004
0.017
<0.004
<0.004
<0.004
<0.002
<0.002
0.017

bag
<0.002
<0.002
0.016
<0.002
<0.002
<0.002
<0.002
<0.002
0.032

<0.002
0.005
<0.002
<0.002
<0.002
0.01
<0.004
<0.002
<0.002
<0.002

-------
FUEL  CELL  TESTING

Fuel Cell  Performance

       The  power plant used  in  the demonstration is a commercial ONSI PC25  200-
kW phosphoric acid fuel cell.  The power plant, designed to operate on pipeline
natural gas, was shipped to and  installed at the Penrose Landfill during 1994.
The field  test began on December 7, 1994, and ended on February 19, 1995,  with
eight  operational test runs.   During this period, the fuel cell operated for
707 hours  on landfill gas.  Power produced by the unit was fed into- the
electrical grid for sale to the  local electrical utility, the Los
Angeles Department of Water and  Power  (LADWP).  It was the first fuel cell ever
connected  to the LADWP grid.   The revenue produced by the sale of electricity
was used to help offset program  costs.

       Because landfill gas  has approximately half the heat content of natural
gas, this  fuel cell power plant  cannot produce 200 kW of net power when
operated on landfill gas.   To  increase the net power, higher flows of landfill
gas would be required to obtain  an equivalent natural gas fuel content and -
heating value.  However, the required modifications to achieve fuel rated  power
were not accomplished for this project.  The only modifications to the fuel
cell were those that could  be  field installed under the direction of an IFC
engineer, from modification kits designed and fabricated by IFC.  The
modifications included a larger  fuel control valve and fuel flow venturi,  a new
process fuel recycle orifice,  a  new cathode exit orifice, and a new redundant
start  fuel shut-off valve.  Additionally, there were modifications to the
control software.

      The fuel cell was operated at up to 137 kW, which is 3 kW below the  goal
for operation on the Penrose Landfill gas.  An-endurance operating condition of
120 kW was selected for the bulk of the field test operations to provide a
margin for steady fuel cell operation during periods of sub-standard gas
quality,  which occur periodically due to upsets in gas quality from the active
landfill (Bradley)  which supplies gas to the Penrose sife.  The fuel cell
efficiency was calculated over two periods during the field test.  The first
period covered 6 days from January 24 through 30, 1995.   Efficiency during this
6-day period of continuous operation was 37.1%.   The second period covered 8
days from February 9 through 17,  1995.  "Average efficiency for this 8-day
period, which included a brief shutdown, was 36.5%.
                                     11

-------
       Table 3 presents results of emission testing of the PC25 power plant at
 the Penrose Landfill  conducted during February 1995.  The emission levels
 indicate that fuel  cells can operate on landfill gas while maintaining the low
 emission levels characteristic of natural-gas-based fuel cells.

                 TABLE 3.   EMISSION TEST* RESULTS FROM FUEL CELL
                            OPERATING ON LANDFILL GAS
                Pollutant
           Fuel Cell Emission
 Sulfur  Dioxide (S02)
 Nitrogen Oxides  (NOX)
 Carbon  Monoxide  (CO)
Non-detect**
0.12 ppm, avg.
0.77 ppm, avg.
 *Tests  used EPA Methods  6c,  7e,  and 10.
 **Detection limit  for  SO2 was 0.23 ppm; all  data  are dry measurements corrected
 to  15%  O,.	    	     .	
GPU Operation

      The GPU consistently removed total halides (as HC1)  from inlet  levels of
45 to 60 ppm in the raw landfill gas  to very low or undetectable levels  at  the
outlet. During seven tests conducted  between January 19  and February  17,  1995,
there were no detectable halides  (individual species detection limits range
from 0.001 to 0.020 ppm).  The results  are consistent with the data contained
in Table 2.

      Total sulfur (as H2S)  was reduced  from about  110 ppm  (about 10 ppm £rom
organic sulfur, plus 100 ppm H2S)  to between non-detectable and 0.385 ppm.  The
only sulfur species detected was carbonyl sulfide.   The  elevated levels  of
0.173 to 0.385 ppm of cabonyl sulfide measured on February 9 and 10,  1995,  are
believed due to a slight increase in  H2S  exiting  the non-regenerable H2S
removal bed, since H2S  at  the H2S bed  exit was measured' at  1.0 to 2.7  ppm on
February 14.  Earlier laboratory work at IFC showed that H2S is  converted to
carbonyl sulfide over the activated alumina in the  downstream drier bed  by  the
reaction   H2S  -t-  C02 = COS + H2O,  due  to  the removal of the product water by the
alumina.  The resulting carbonyl sulfide is not readily  removed by the low
temperature carbon bed.  The non-regenerable H2S  removal bed was  switched over
to a fresh bed on February 15, 1995,  and the exit H2S level returned to non-
detectable.  The carbonyl sulfide level  measured shortly after,  on February 17,
also fell to just 0.061 ppm.
                                   12

-------
DISCUSSION  AND   CONCLUSIONS

      Emissions from landfills are potential contributors to global warming and
hazardous air pollutants.  Conventional methods to mitigate these emissions,
such as flaring, produce other greenhouse gases, such as CO2.  Operating a fuel
cell at a landfill site can eliminate CH4,  hazardous air pollutants,and other
universal secondary emissions  (CO, SO2,  NOX) ,  lowers  total CO2 emissions, and
can permit efficient generation of electric power.  In order to operate a fuel
cell on landfill gas, the gas roust be sufficiently purified.  A landfill gas
cleanup pretreatment module designed, constructed, and tested in this
demonstration was successful.  The combined gas cleanup system and power plant
produced electrical power with low levels of air pollution.

      Currently, the fuel cell/GPU is located on a landfill in Groton,
Connecticut, as a follow-up to the earlier demonstration of the technology at
the Penrose Landfill.  This demonstration is scheduled to run for approximately
18 months during which time a significant amount of operating data will be
acquired.   Additionally, research will be conducted to refine the design of the
GPU in order to simplify the required equipment.  This demonstration is a
partnership between EPA, the Town of Groton, and Connecticut Light and Power,
with technical support from IFC.

      It is expected that at least 1.6 million kWh of electricity will be
generated over the test period.  The fuel cell has produced a record 165 kW
from the Groton Landfill gas, but the system is being operated at a
conservative 140 kW to minimize inadvertent shutdown due to gas heat
fluctuations.
                                   13

-------
REFERENCES

[1]   Air Emissions from Municipal Solid Waste  Landfills -  Background
      Information for Proposed Standards and Guidelines (EPA-450/3-90-011a;-
      NTIS PB91-197061), March 1991.

[2]   Sandelli, G. J., Demonstration of Fuel Cells  to Recover  Energy from
      Landfill Gas, Phase I Final Report: Conceptual  Study  (EPA-600/R-92-007;
      NTIS PB92-137520), January 1992.

[3]   Trocciola, J. C. and Preston, J. L., Demonstration of Fuel Cells to
      Recover Energy from Landfill Gas, Phase II, Pretreatment System
      Performance Measurement (EPA-600/R-95-155; NTIS PB96-103601), October
      1995.

[4]   Graham. J. R. and Ramaratnam, M., Recovery of VOCs Using Activated
      Carbon, Chemical Engineering, pp. 6-12, February 1993.
                                       14

-------

-------

-------