Fuel Cell Report

FEASIBILITY STUDY FOR FUEL CELL RURAL ELECTRIFICATION IN SCATTERED AND CLUSTERED COMMUNITIES

Final Report for Energy Systems II Professor: Dr. Barriga Submitted by: Stephen Welty University of Calgary

Table of Contents Table of Contents................................................................................................................ 2 List of Figures .................................................................................................................... 2 Introduction ........................................................................................................................ 3 Fuel Cells............................................................................................................................ 3 Fuel cell operation ......................................................................................................... 3 Efficiencies ..................................................................................................................... 5 Applications ................................................................................................................... 6 Fuels and Methods ............................................................................................................. 7 Natural Gas.................................................................................................................... 7 Other hydrocarbons...................................................................................................... 8 Hydrogen........................................................................................................................ 8 Life cycle analysis considerations...................................................................................... 8 Economic Analysis ........................................................................................................... 10 Establishing the demand cases ................................................................................... 10 Stand alone power generation.................................................................................... 10 Clustered Community ............................................................................................... 10 Local production of fuels for fuel cells ...................................................................... 15 Waste Heat and Exhaust Recovery.................................................................................. 18 Conclusions ...................................................................................................................... 19 Bibliography ..................................................................................................................... 21

List of Figures Figure 1: Schematic for the operation of PEM and Phosphoric Acid fuel cells................ 4 Figure 2: Comparison of efficiencies for ideal heat engines and Fuel Cells with a Carnot low temperature of 30C....................................................................................................... 6 Figure 3: Greenhouse Gas emissions of energy options. ................................................... 9 Figure 4: The effect on UCE for different values of installed costs for the SOFC system 15 Figure 5: Waste heat and water exhaust for different energy generation levels. ............ 18

Stephen Welty

Page 3

5/26/2002

Introduction Currently, there are roughly 2 billion inhabitants of this planet that do not have access to electricity. In the mind of modern western man, electricity is a basic need and the lack thereof implies a significantly lower quality of life. To accommodate those of the roughly 2 billion people without electricity that would like to gain access to the western way of life, a number of rural electrification schemes have been devised both from the private sector and the government. The most common way to provide inhabitants with electricity is with the electricity grid. However, this is not always possible or economically feasible and other distributed energy generation methods have been employed. This paper is aimed at evaluating a relatively new technology for this application. This study has as its main goal to evaluate different fuel cell technologies that would be suitable for electricity generation in small off-grid communities. The study will consider both scattered settlements and clustered settlements where two different power generation schemes will have to be considered to be economically feasible. Fuel cells can use available hydrocarbons ranging from methane to diesel but different methods will have to be used for different fuel supply scenarios and these will be discussed. Additionally, fuel cells can be combined with renewable energy that provides irregular supply of energy. In this case energy would be stored in hydrogen through electrolysis and that hydrogen would be supplied to the fuel cell to meet electricity demand. Essentially, the proposition is that fuel cells could replace batteries in renewable energy systems if it is economically viable. An important possible advantage of fuel cells is supplying clean drinking water for the community. A number of fuel cells have wastewater exiting the system as steam, which could be condensed and used as relatively pure drinking water. This possibility will depend on the fuel cell technology used and the purity of the fuel supply. In this study a simple economic analysis tool was applied to give a first order view of the economics of the fuel cell generating system compared with other available generating systems. A brief view of the life cycle analysis of fuel cells will be discussed to determine its real environmental impacts compared with other technologies as well as to determine its economic feasibility in terms of reliability, durability and replacement costs.

Fuel Cells Fuel cell operation A fuel cell is an electrochemical device, which produces DC electricity directly from the chemical energy of a substance (usually hydrogen). It is very similar to a battery with the main difference being that it is an open system with the fuel constantly supplied to the

Stephen Welty

Page 4

5/26/2002

device. Fuel cells are comprised of two electrodes, an electrolyte and an external circuit. The electrolyte provides for ion migration from one electrode to the other. One of the electrodes is a positive anode and the other is a negative cathode. The reactants, hydrogen in most cases, lose electrons at the negative electrode through an oxidation reaction and the electrons travel through an external circuit to the positive electrode. The positively charged hydrogen ions from the oxidation reaction travel through the ionic conductor or electrolyte to the positive electrode. At the positive electrode a reduction reaction takes place where the electrons combine with the positively charged hydrogen ions and oxygen to form exhaust water. In the process, heat is also released which in some applications can be captured for useful thermal energy. This process is typical of Proton Exchange Membrane and Phosphoric Acid fuel cells but there are other processes using different fuels and in some cases the oxygen picks up electrons at the cathode and then migrates to the anode to react with the positive hydrogen ions.

Figure 1: Schematic for the operation of PEM and Phosphoric Acid fuel cells1.

There are a number of different methods to produce the effects described above leading to a number of different technologies within the fuel cell technology general field. Table 1 is a list of the different fuel cell types and some of their characteristics. The list is not comprehensive and there are a number of differences between manufacturers materials and methods for manufacturing fuel cells that in many cases are proprietary and confidential. The types of fuel cells can be broken down into two different categories: low temperature fuel cells and high temperature fuel cells. The low temperature fuel cells include AFC, PEM and PAFC and the high temperature technologies are MCFC and SOFC. The newest technologies are the DMFC and URFC, which are furthest from commercialization. PAFC is the most commercially developed fuel cell type and has been used most in stationary power supply. The PEM or SPFC technology is the most promising technology for transport applications since it has a high power density and a quick response time to load but it is also suitable for stationary power applications and when it is commercialized, it is expected to be the lowest cost fuel cell technology. AFC technology was the first developed fuel cell technology and was used in space 1

Taken from americanhistory.si.edu website.

Stephen Welty

Page 5

5/26/2002

applications in the 1960’s but the costs of this technology have been considered to be too high for commercialization. Therefore, it has received little attention from manufacturers and has received little funding for development. SOFC span the largest application range and are considered slated for commercialization before 2004. They also could have the highest efficiencies of the fuel cell technologies when combined with gas turbine energy generation (up to 70%). A single cell (a combination of two electrodes and an electrolyte) produces only about 0.7 volts under load. In order to achieve a practical voltage, the cells are put together into a “stack”. To do this, an interconnect is put between the anode of one cell and the cathode of the next cell. Fuel Cell Type

Acronym

Operating Temp [ºC] 150-200

Achieved Efficiencies 40%

Electrolyte

Electrodes

Fuels

Alkaline

AFC

KOH in H2O

PEM or SPFC

80-100

35 to 40%

Solid Polymer

Platinum catalysts Platinum catalysts

H2 and O2 Compressed Purified H2

Proton Exchange Membrane Phosphoric Acid Molten Carbonate Solid Oxide

PAFC

150-200

35 to 40%

MCFC

650

50 to 55%

Platinum catalysts Ni catalysts

Gasoline, H2, Natural Gas H2

SOFC

1000

45 to 50%

Phosphoric Acid Na, K or Mg Carbonates Ca or Zi oxides

H2 No Reformer

DMFC

80-100

35 to 40%

Solid Polymer

La-Mn cathode and Ni-Zr anode Pt and Ru

Direct Methanol Regenerative

Methanol

URFC

80-100

35 to 40%

Solid Polymer

Varies

Purified H2

Table 1: Fuel cell types and their characteristics. Efficiencies based on electricity out/Lower heating value of fuel in.

Efficiencies The efficiencies of the fuel cells listed in table 1 range from 35% to 55%. To get a better understanding of the efficiencies it is useful to compare these efficiencies to the maximum possible efficiencies from internal and external combustion machines. Fuel cells are not subject to the thermodynamic efficiencies associated with the second law of thermodynamics and the Carnot efficiency limitation. The Carnot efficiency is defined for a heat engine whose temperature extremes are known as follows: T − TL Carnot = H TH where TH is the high temperature reservoir and TL is the low temperature reservoir. The theoretical maximum efficiency of an electrochemical device such as the fuel cell is defined not by its temperature extremes but by the ratio of the Gibbs free energy value (∆Gº) over the enthalpy value (∆Hº) or total heat energy of the fuel at a given temperature. Figure 2 is a graph comparing the two theoretical efficiencies assuming a 30ºC low temperature for the heat engines. The discontinuity in the graph of the fuel cell efficiency at 100ºC is due to the change in phase of the exhaust water from liquid to gas.

Stephen Welty

Page 6

5/26/2002

This change in phase affects both the enthalpy of the water and the entropy, which directly affects the Gibbs free energy value. An important aspect of fuel cells is their low pollution. For the low temperature types, it is critical to have a relatively pure supply of hydrogen and this results in pure exhaust water with no pollutants such as NOx or SOx or CO2. However, in some cases the method of extracting hydrogen from a hydrocarbon may release some pollutants but on a much smaller scale than combustion of those fuels. With the high temperature fuel cells the emissions are also very low though some SOx are produced and in the case of SOFC with an operating temperature of 1000ºC, there is the possibility of forming NOx. If hydrogen is obtained from electrolysis of water using renewable energies, then there are no pollutants released because of the pure hydrogen. Additionally, there would be no carbon dioxide emissions. 100.00%

Fuel Cell

90.00% 80.00%

Discontinuit y due t o change from liquid t o gas phase of exhaust wat er (Delt a Enthalpy = 41kJ/ mol@100C)

Efficiency %

70.00% 60.00% 50.00% 40.00% 30.00%

20.00%

Carnot Heat Engine

10.00% 0.00% 0

200

400

600

800

1000

1200

Tem perature [C]

Figure 2: Comparison of efficiencies for ideal heat engines and Fuel Cells with a Carnot low temperature of 30ºC

Applications The applications for fuel cells are quite broad and rather specific to fuel cell type. The applications range from replacement of batteries in consumer electronics to large-scale electricity generating plants. Table 2 gives an idea of the applications and reasonable power ranges for a module of the different fuel cell types. The applications of interest for this study are domestic power and small-scale power. It can be seen from the table that all of the fuel cell types would be applicable to the case of a small clustered community requiring small-scale power and most of the technologies would be applicable for domestic power. MCFC technology would not be applicable to domestic power because the minimum module size is reported to be 250kW, which is

Stephen Welty

Page 7

5/26/2002

much more than an individual household would need. The only technologies available for a 1kW domestic power fuel cell are AFC, SPFC and SOFC. Fuel cell type AFC SPFC PAFC MCFC SOFC

Module size range [kW] <1-200 <1-500 5-500 250-5000 1-5000

Waste heat output [ºC] <60 <80 <200 600 850

Waste water output liquid liquid Steam Steam Steam

Domestic power

! ! " " !

Small-scale power

Large-scale cogeneration

Trans port

Battery Replacement

" " " ! !

! !

" ! " "

! ! ! ! !

-

" "

-

2

Table 2: Applications of different fuel cell types .

Fuels and Methods Natural Gas Natural gas can be obtained in a number of ways including extraction in associated and non-associated gas fields of fossil fuels and production from biomass. If the gas is extracted from fossil fields, then it cannot be considered a renewable energy despite the fact that the combustion or electrochemical process of this fuel produces much less carbon dioxide than other hydrocarbons. However, as shown in figure 3, there are possibilities for sequestering the carbon dioxide to avoid greenhouse gas emissions. The most environmentally friendly way of using natural gas is to produce it from biomass through thermal gasification or biological conversion known as biogas.

Figure 3: Sequestering carbon dioxide in a fossil fuel powered fuel cell3.

In order to use natural gas in low temperature fuel cells, a reformer must be used which separates the hydrogen from the carbon. The most common reforming method is steam reforming which operates in two steps. In the first step the natural gas is exposed to high 2

Adapted from “An introduction to fuel cell technology and economics” by Nigel Brandon and David Hart.

3

Taken from http://www.fuel-cell.de website.

Stephen Welty

Page 8

5/26/2002

temperature steam and broken down into hydrogen, carbon monoxide and carbon dioxide. The second step consists of exposing the carbon monoxide from the first step to high temperature steam and producing more hydrogen and carbon dioxide. This process can reach efficiencies of 70 to 90%. The hydrogen and carbon dioxide are then sequestered and stored in tanks. The hydrogen can then be used directly as a reactant for the fuel cell. In high temperature fuel cells there is a possibility of internal reforming of the natural gas in the fuel cell, which would not require an extra reformer. The issues for internal reforming are the same as the external reforming case.

Other hydrocarbons Hydrocarbons such as gasoline, coal, LPG and methanol can also be used as fuels in fuel cells. These other hydrocarbons still require a reforming process that will release even more carbon dioxide than the reforming of methane and reduce the overall efficiency of the system. The only advantage to using other hydrocarbons in fuel cells is the fuel availability. The most desired hydrocarbons for hydrogen production are those fuels that have the highest hydrogen to carbon ratio and the least amount of other components. Natural gas is the best hydrocarbon for the production of hydrogen.

Hydrogen Since hydrogen is the reactant used in most fuel cells it would be ideal to get the fuel supplied directly as hydrogen. If pure hydrogen is supplied to the fuel cell the only products will be water and heat. This would lead to a zero-emissions technology for generating electricity, which would be very attractive in light of current environmental concerns. The difficulty with supplying pure hydrogen to the fuel cell is its availability and difficulties in storage. It could be obtained by electrolysis of water to obtain oxygen and hydrogen but this process requires energy. Some advocates of this method suggest using renewable energies to provide the energy for the electrolysis process. This would reduce the problems associated with the intermittent nature of the availability of renewable energy technologies such as photovoltaics, wind energy and micro hydro. In this context, the hydrogen becomes an energy storage and transport method rather than a fuel as such. There are other methods of obtaining hydrogen such as thermal water splitting, thermochemical cycles which use high temperature heat to split water, photoelectrolysis (sunlight directly splits water into hydrogen and oxygen), photobiology (using sunlight and microorganisms), and radiolysis. Although these methods may have future potential, they are currently a long way from commercialization.

Life cycle analysis considerations Life cycle analysis is important in determining the economics, environmental impacts and general feasibility of a technology. Life cycle analysis is also known as “cradle to grave” assessment of a technology. In the case of fuel cells this includes the energy to fabricate the device, the energy to produce hydrogen fuel, the carbon dioxide and other emissions in the process of removing hydrogen from hydrocarbons, and the energy required to recycle or dispose of the device at the end of its useful life. This life cycle analysis is

Stephen Welty

Page 9

5/26/2002

used to determine the environmental impact of a technology but some of these issues can be translated into dollar costs for an economic analysis.

Figure 4: Greenhouse Gas emissions of energy options4.

Figure 4 shows the greenhouse gas emissions from given technologies per TWh of energy produced. The lowest emitting technologies are hydro, wind, solar and nuclear. The emissions of fuel cell technology is calculated based on a natural gas supply and are similar to the emissions of a natural gas power plant. However, as discussed in the previous section, other sources of hydrogen could be used which would reduce the emissions of fuel cells to a level similar to wind and solar. An important aspect in the life cycle analysis of a technology is the life of a device. For instance, if an electrification program were to last 20 years and the device had a projected life of 5 years, the devices would have to be replaced 4 times in the duration of the project which would add to the cost of the project and to the environmental impact. For PAFC’s, the most commercialized fuel cell technology, the life is projected to be about 40,000 hours or five years. The short life of PAFC technology seems to be characteristic of technologies using liquid electrolytes. The MCFC technology also uses liquid electrolytes and has a similar life expectancy. However, SOFC’s, which use a solid ceramic as their electrolyte, are expected to have lifetimes of 10 to 20 years. PEM fuel cells also use a solid electrolyte, which would yield longer lifetimes but they are “poisoned” by gases from impure fuel. It is necessary to have extensive fuel processing

4

Taken from “Comparing power generation options” by Hydro Quebec.

Stephen Welty

Page 10

5/26/2002

facilities before the fuel can be used in the fuel cell, which both increases price and reduces efficiency down to about 42%. At this time there is not enough information on manufacturing, disposal or recycling costs for fuel cell technologies to do an in depth life cycle analysis. It is sufficient to say that the operational emissions and energy consumption is much less than diesel generators and coupled with renewable energies could offer near zero operational emissions.

Economic Analysis Establishing the demand cases For the purposes of doing a preliminary economic analysis of fuel cells, an arbitrary community has been devised. For the case of the clustered and scattered communities there are 800 households each consuming 800Wh/day leading to a total energy demand of roughly 640 kWh/D. For the clustered community it was determined to be more economical to get a central generation station and transmit the electricity to the different households via transmission wires. It is assumed that the transmission lines are already in place due to a previous generation station that now needs replaced. For the scattered community, there are no transmission lines and it is considered more economical to generate power at each individual household. Item Quantity Load [W] Use [h/d] Energy [Wh/d] Lights 4 13 5 260 Radio 1 20 3 60 TV 1 30 3 90 Refrigerator 1 16 24 384 Per family Peak 55.3 Daily 794 Community Peak 28756 Daily 635200 Table 3: Distribution of Consumption for a typical household. Refrigerator consumption based on an efficient single temperature refrigerator.

Table 3 shows the consumption of a typical household. The per family peak demand is calculated assuming only 70% of the items will be on at the same time during the peak hours of demand for the community. The community peak demand is calculated assuming a 0.65 simultaneity factor. The peak demand that the generation system must be designed for is about 29kW.

Stand alone power generation Clustered Community The first case to be studied will be the clustered community where the generation system will be put in a fixed location and transmission wires will supply the community with electricity at the point of use. As mentioned earlier this case will be studied assuming that there was a diesel generator at the location previously and that the transmission lines are already in place. This assumption gives an advantage to the central generation scheme since putting in transmission wires could be a significant cost depending on the layout of the community. However, the object of this study is not to compare central

Stephen Welty

Page 11

5/26/2002

generation to distributed-generation but to compare different technologies within a given scheme of electrification. Table 4 gives the values for the costs associated with putting in another diesel generator. Total Energy Demand per day Peak demand for the community Heating Value of Diesel

29 kW 42000 kJ/kg

Gen-Set with efficiency of: Gen-set $/kW system cost: Safety factor for peak demand: Interest rate for capital

15.00% $400 $/kW 1.3 12.00%

Conversion kg/gal of diesel Regular overhaul every 5 years

Year

640 kWh/D

3.30 kg/gal

Required capacity of gen-set

37.7 kW

Price for this gen-set

$15,080

Life of the gen-set

170,000 hours

Maintenance/year is assumed 3% of I.C.

$452 $/year

Labor per year with one person 1/4 time:

$2,080 $/year

Other unforeseen costs:

$1,000 $/year

Project Duration

20 years

Diesel

$1.20 USD/gal

$3,000

Present Value of overhaul Fuel Consumption/year (345 days)

Total Initial Cost

5

$1,702.28

2304000 kJ/D

10

$965.92

365.71 kg/D

Yearly fuel cost

15

$548.09

126171.43 kg/year

Total yearly cost

$18,296 USD

Annuity

$2,449 USD/year $45,881 USD/year $51,862 USD/year UCE

$0.23 USD/kWh

Table 4: Unified Cost of Electricity for a Gen-Set in a clustered community with existing transmission lines.

The yearly fuel cost for this generation scheme is about 88% of the total yearly cost, therefore, any fluctuation in diesel price during the duration of this project could have a significant impact on the UCE. Another case for electrification of this community is shown in Table 5. A Phosphoric Acid fuel cell was selected for this case because it is the only commercially available fuel cell to date. The PC25 is a PAFC manufactured by United Technologies Company and has a capacity of 200kW. The fuel cell needed in this case is only a 29kW fuel cell. Although that model is not currently available, the analysis was done assuming that when such a model is available, it would cost roughly the same in $/kW as the PC25. However, as these fuel cells are manufactured in larger volumes, the prices will certainly come down and different technologies will fit into different size categories. It may be the case that a 30kW PAFC is never available but it is very likely that a SOFC or PEMFC will be available in the near future to accommodate this size range. It is also important to note that the expected lifetime of a PAFC is only about 40,000 hours and will require replacement in about the same intervals as batteries in a photovoltaic system. SOFC and PEMFC using solid electrolytes have the potential to last much longer. The calculations shown in table 5 include the cost of replacement every 5 years and even with this, a PAFC at 3,500$/kW has a similar unified cost of electricity as a diesel generator. It is also important to note that the methane cost was taken at $250/thousand cubic meters ($0.35/kg), which was the cost of natural gas in the United States in 2000. In Ecuador, for example the cost of LPG, which could also be reformed to extract the hydrogen for use in fuel cells, is 0.10$/kg. The yearly cost of fuel in the calculations in table 5 is only 23% of the total yearly cost making this system much less susceptible to the volatility of fossil fuel prices. Additionally, methane can be produced through thermal gasification of

Stephen Welty

Page 12

5/26/2002

biomass or in biogas digesters, which make it possible for fuel cells to be a renewable energy and assure the security of supply for remote regions. Total Energy Demand per day Peak demand for the community Heating Value of Methane PAFC with efficiency of: PAFC $/kW system cost: Safety factor for peak demand: Interest rate for capital

640 kWh/D 29 kW 54000 kJ/kg 40.00% $3,500 $/kW 1.3 12.00%

Density of Methane

0.72 kg/m^3

Required capacity of fuel cell

37.7 kW

Price for this fuel cell

$131,950

Life of the fuel cell

40,000 hours

Maintenance/year is assumed 3% of I.C.

$3,959 $/year

Labor per year with one person 1/4 time:

$2,080 $/year

Other unforeseen costs:

$1,000 $/year

Project Duration

20 years

Methane costs

$0.25 USD/m^3

Inverter Year

PV of Replacement

Fuel Consumption/year (345 days)

5

$74,871.97

2304000 kJ/D

10

$42,484.37

15

$24,106.77

$5,000.00 USD Total Initial Cost

$278,413 USD

Annuity

$37,274 USD/year

106.67 kg/D

Yearly fuel cost

$12,778 USD/year

36800 kg/year

Total yearly cost

$57,090 USD/year

UCE

$0.26 USD/kWh

Table 5: Generation scheme using a PAFC at current market value using existing transmission lines.

The case shown in table 6 is a little more future-based than the previous two cases since there are no commercially available SOFC to date. The reason for doing this case was to investigate the effect on the UCE if instead of being replaced every 5 years, as the PAFC Total Energy Demand per day Peak demand for the community Heating Value of Methane SOFC with efficiency of: SOFC $/kW system cost: Safety factor for peak demand: Interest rate for capital

29 kW 54000 kJ/kg 50.00% $2,000 $/kW 1.3 12.00%

Density of Methane

0.72 kg/m^3

Regular overhaul every 5 years Year

640 kWh/D

$5,000

Required capacity of fuel cell

$75,400

Life of the fuel cell

170,000 hours

Maintenance/year is assumed 3% of I.C.

$2,837.13

2304000 kJ/D

10

$1,609.87

15

$913.48

$2,262 $/year

Labor per year with one person 1/4 time:

$2,080 $/year

Other unforeseen costs:

$1,000 $/year

Project Duration

20 years

Methane costs

$0.25 USD/m^3

Inverter

Present Value of overhaul Fuel Consumption/year (345 days) 5

37.7 kW

Price for this fuel cell

$5,000.00 USD Total Initial Cost

$85,760 USD

Annuity

$11,482 USD/year

85.33 kg/D

Yearly fuel cost

$10,222 USD/year

29440 kg/year

Total yearly cost UCE

$27,046 USD/year $0.12 USD/kWh

Table 6: Unified Cost of Electricity for the case of a SOFC in the clustered community

would have to be, there was just an overhaul cost every five years. The overhaul cost is taken as 6% of the initial cost of the equipment every five years. It is uncertain how accurate this value is since there is no field experience available other than some field trials, which were not available to the author for inclusion in this paper. It is reasonable to assume, however, that the overhaul and maintenance cost of fuel cells will be much lower than diesel generators since there are no moving parts in a fuel cell. SOFC’s also have the potential to be from 50% to 55% efficient, reducing the fuel consumption of the system. The UCE for this case is 0.12$/kWh which is getting competitive with largescale power plants supplying the grid. And this is based on a price of $2000/kW, which

Stephen Welty

Page 13

5/26/2002

is predicted by some companies to drop to $800/kW in volume production. The only way of verifying the veracity of these statements is to wait until the technology is available. The last case investigated for the clustered community was the use of a traditional photovoltaic system using batteries. Table 7 shows the values used in the calculations for this case and the results. Total Energy Demand per day Peak demand for the community Daily Solar Energy

640 kWh/D 29 kW

Autonomy of batteries Depth of Discharge

4.2 kWh/m^2 Battery Safety Factor

D.S.E. Factor

0.65

Required capacity of batteries

PV system efficiency

10%

Battery price/kwh

1.3

Price for batteries

Safety factor for peak demand: Peak Wattage

234432 W(p)

PV system cost:

$6 $/W

Cost of PV panels Required Area

$1,406,400 USD 2344.32 m^2

Project Duration

20 years

Present Value of battery replacement

1.3 4160 kWh $165 $/kWh $686,400 USD

Life of the batteries

40,000 hours

Maintenance/year is assumed 5% of initial cost Labor per year with one person 1/30 time: Other unforeseen costs: Interest rate for capital Inverter

Year

3.00 days 60.00%

$8 $/year $277 $/year $20 $/year 12.00% $5,000.00 USD

Total Initial Cost

$2,833,687 USD

5

$389,481.79

Annuity

$379,371 USD/year

10

$221,002.43

Total yearly cost

$379,676 USD/year

15

$125,402.71

UCE

$1.72 USD/kWh

Table 7: Photovoltaics with batteries for the clustered community with existing transmission lines.

It is clear from the results of this case that solar electricity is the least attractive technology economically unless fuel supply is tremendously difficult. The UCE is more than five times higher than the most expensive alternative analyzed. The main advantages of solar power have traditionally been that no fossil fuel is required, and that it is an environmentally friendly technology. These studies suggest that in the case of small-scale centralized electricity generation, fuel cells may fill the niche that photovoltaics began to fill and may in many cases displace diesel generators. Scattered Community The case of the scattered community requires onsite power generation at each household. Traditionally this has been done with the use of photovoltaic panels and possibly wind turbines. The difficulty with using diesel generators for on-site generation is the noise, pollution and the access to fuel since many times these dwellings are in difficult to reach areas. Table 8 shows the case for a photovoltaic system for each household. In the central generation schemes analyzed for the clustered community it is necessary to have a significant amount of labor but since the worker is working to generate much more power than the cases of the scattered community, the fraction of labor in the overall cost will be much lower. In all of the cases analyzed for the scattered community, labor was included, but it can be argued that with sufficient training the system can be maintained by the homeowner with a small amount of his time. This would yield the labor costs

Stephen Welty

Page 14

5/26/2002

irrelevant. With labor costs in table 8 the UCE is $2.86/kWh and without labor costs it would be $1.85/kWh. Per family Energy Demand per day. Peak demand of the household Daily Solar Energy

0.8 kWh/D 0.055 kW

Autonomy of batteries

3.00 days

Depth of Discharge

60.00%

4.2 kWh/m^2 Battery Safety Factor

1.3

D.S.E. Factor

0.65

Required capacity of batteries

PV system efficiency

5.2 kWh

10%

Battery price/kwh

$165 $/kWh

Safety factor for peak demand:

1.3

Price for batteries

$858 USD

Peak Wattage

293 W(p)

PV system cost:

$6 $/W

Cost of PV panels

$1,800 USD

Required Area

2.93 m^2

Project Duration

Life of the batteries

40,000 hours

Maintenance/year is assumed 5% of I.C.

$8 $/year

Labor per year with one person 1/30 time:

$277 $/year

Other unforeseen costs:

20 years

$20 $/year

Interest rate for capital

12.00%

Inverter Year

$30.00 USD

Present Value of battery replacement

Total Initial Cost

$3,608 USD

5

$486.85

Annuity

$483 USD/year

10

$276.25

Total yearly cost

$789 USD/year

15

$156.75

UCE

$2.86 USD/kWh

Table 8: Unified Cost of electricity for one household in a scattered community using a conventional PV system.

Table 9 shows the case for a small-scale fuel cell in the scattered community. This case also includes labor costs but the same argument could apply to this system as to the photovoltaic system. With labor the UCE is $1.34/kWh whereas without the labor costs the UCE would be $0.33/kWh. Per family Energy Demand per day.

0.8 kWh/D

Peak demand of the household

0.055 kW

Heating Value of Methane

54000 kJ/kg

SOFC fuel cell with efficiency of: Take SOFC to have $/kW cost of: Safety factor for peak demand: Interest rate for capital

$2,000 $/kW 1.3 12.00%

Density of Methane Regular overhaul every 5 years Year

50%

Present Value of overhaul

Required capacity of fuel cell

0.0715 kW

Price for this fuel cell

$143

Life of the fuel cell

170,000 hours

Maintenance/year is assumed 5% of I.C.

$7 $/year

Labor per year with one person 1/30 time:

$277 $/year

Other unforeseen costs:

$20 $/year

Project Duration

0.72 kg/m^3

Methane costs

$200

Inverter

Fuel Consumption/year (345 days)

20 years $0.25 USD/m^3 $30.00 USD Total Initial Cost

$387 USD

5

$113.49

2880 kJ/D

Annuity

$52 USD/year

10

$64.39

0.11 kg/D

Yearly fuel cost

$13 USD/year

15

$36.54

36.8 kg/year

Total yearly cost

$369 USD/year

UCE

$1.34 USD/kWh

Table 9: Unified cost of electricity for one household in a scattered community using a SOFC.

This case was done using a SOFC cost of $2000/kW and assuming that the lifetime would be 170,000 hours and the system only requires an overhaul every five years. Since the SOFC cost is not known and there are many claims about the potential of reducing the costs in volume production, it is instructive to investigate the effects on the unified cost of electricity for different price ranges. Figure 5 shows the effect on the

Stephen Welty

Page 15

5/26/2002

UCE for changing values of the installed cost of the SOFC system without labor costs. The graph shows that even for relatively expensive SOFC systems, the UCE is much lower than that of the photovoltaic system without labor ($1.85/kWh).

$0.60

UCE $/kWh

$0.50 $0.40 $0.30 $0.20 $0.10 $0.00 $0

$1,000

$2,000

$3,000

$4,000

$5,000

$6,000

Installed Cost $/kW

Figure 5: The effect on UCE for different values of installed costs for the SOFC system

Even in the on-site generation scheme investigated here, the fuel cell technology has a clear economic advantage over photovoltaics. There are some assumptions made in the analysis that should be emphasized again. It is assumed that fuel cell technology will be available in these size ranges in the near future and that the lifetimes will be 170,000 hours or more so that frequent replacement is not necessary.

Local production of fuels for fuel cells One possible disadvantage to using fuel cells in remote communities is the need for a constant supply of fuel as in the case of the diesel generator even though the fuel required is much less for fuel cells. However, a number of local hydrogen or methane generation methods are possible for fuel cells. Among them are biogas from digesters, producer gas from thermal gasification of biomass and electrolysis of water using renewable energies. For the biogas digesters, approximately 0.5-0.6 m3 of biogas can be produced per kilogram of volatile solids added to the digester. Biogas is composed mainly of methane (60%) and carbon dioxide (40%) with small amounts of other gases. While it is not the aim of this paper to do an in depth analysis of the biogas potential, it is important to note that it is technically and economically feasible. A 3 m3 digester will produce roughly 1 m3 of biogas per day under optimal conditions. Since the digester can be constructed of local materials and local labor, it becomes an economically feasible endeavor. Additionally, the added social and health benefits must be considered. A properly designed and maintained digester will eliminate over 90% of the disease causing agents from animal and human wastes. In areas where using human waste directly as fertilizer is a common practice there wouldn’t be much of a cultural barrier to using human waste in the digester and it would considerably increase the quality of their foods. In areas where

Stephen Welty

Page 16

5/26/2002

human waste is not traditionally used as fertilizer, there may be some resistance to using this technique. For the case of the SOFC in table 6, approximately 120 m3 of methane is required which translates to about 200 m3 of biogas at 60% methane content. This biogas supply could be met by a community bio-digester of 600 m3 and an input of 400 kg of volatile solids per day. For the on-site generation SOFC in table 8, a biogas supply of 0.26 m3 is required which could be supplied by a 0.8 m3 household bio-digester with a volatile solids input of 0.5kg per day. However, the technological complexity of a biogas digester is rather high when one considers all of the conditions that must be met for optimal biogas production. Table 10 shows an analysis of using hydrogen produced from photovoltaics to run the SOFC. An electrolysis efficiency of 80% is used along with an electrolysis device cost of 600$/kW. This case is the same as the case in table 7 for the clustered community with the difference of using hydrogen to store the energy and fuel cells to recuperate it when needed. The unit cost of electricity for the conventional PV system using batteries in table 7 is 1.72$/kWh whereas with the system in table 10 the UCE is $2.25/kWh. Total Energy Demand per day

640 kWh/D

Fuel Cell Safety Factor

Peak demand for the community

29 kW

Fuel Cell Capacity

Daily Solar Energy

4.2 kWh/m^2 Fuel Cell Price $/kWh

D.S.E. Factor

0.65

Price for fuel cell

PV panel efficiency

13%

Life of the fuel cell

Safety factor for peak demand: Peak Wattage PV system cost:

170,000 hours

586081 W(p)

Labor per year with one person 1/4 time:

$2,080 $/year

Other unforeseen costs:

$1,000 $/year

System Overhaul every 5 years

$5,000 $/year

4508.31 m^2

Project Duration

$113,100 USD

Maintenance/year is assumed 3% of I.C.

$3,516,600 USD

Required Area

20 years

Inverter

80.00%

Electrolysis Device

Fuel Cell Efficiency

50.00%

Hydrogen Storage Tank

Present Value of battery replacement

$3,393 $/year

$5,000 USD

Interest rate for capital

Electrolysis Efficiency

Year

$3,000 $/kW

1.3

$6 $/W

Cost of PV panels

1.3 37.7 kW

12.00% $16,000.00 USD

Total Initial Cost

$12,000.00 USD $3,668,060 USD

5

$2,837.13

Annuity

$491,075 USD/year

10

$1,609.87

Total yearly cost

$497,548 USD/year

15

$913.48

UCE

$2.25 USD/kWh

Table 10: PV/fuel cell system using hydrogen as energy storage for the clustered community.

The main disadvantage of the PV/fuel cell system is that the efficiency of performing electrolysis coupled with the fuel cell efficiency makes the photovoltaic panels needed much greater. This fact is the greatest contributor to the higher electricity cost of the system. It is likely that a system with wind turbines or micro-hydro to perform electrolysis would be much more economically feasible than a system with photovoltaics. The case for the scattered community is shown in table 11. Once again the main economic disadvantage is the requirement of more PV panels. When compared to the case in table 8 of a conventional PV system with the same demand case, it can be seen that using fuel cells in this context yields no economic advantage increasing the UCE from $2.86/kWh to $3.79/kWh.

Stephen Welty

Page 17

Per family Energy Demand per day. Peak demand of the household

0.8 kWh/D 0.055 kW

Daily Solar Energy

5/26/2002

Fuel Cell Safety Factor

1.3

Required capacity of fuel Cell

4.2 kWh/m^2 Fuel cell price/kw

D.S.E. Factor

0.65

Price for fuel cell

PV panel efficiency

13%

Life of the fuel cell

0.0715 kW $2,000 $/kW $143 USD 170,000 hours

Safety factor for peak demand:

1.3

Maintenance/year is assumed 5% of I.C.

$100 $/year

Peak Wattage

733 W(p)

Labor per year with one person 1/30 time:

$277 $/year

PV panel cost:

$6 $/W

Cost of PV panels

$4,200 USD

Required Area

5.64 m^2

Project Duration

20 years

Other unforeseen costs:

Inverter

$200.00 USD

Electrolysis Device

Fuel Cell Efficiency

50.00%

Electrolysis Efficiency

Present Value of 5 year overhaul

$200 $/year $30.00 USD

Interest rate for capital

Hydrogen Storage Tank

Year

$20 $/year

Regular overhaul every 5 years

12.00% $50.00 USD 80.00% Total Initial Cost

5

$113.49

10

$64.39

Annuity Total yearly cost

15

$36.54

UCE

$4,837 USD $648 USD/year $1,045 USD/year $3.79 USD/kWh

Table 11: Scattered community PV/fuel cell system with hydrogen as energy storage.

The case for thermal gasification of biomass has more potential for being economically feasible than the photovoltaics system of generating hydrogen. In order for producer gas technology to be sustainable, the biomass used in the production must be harvested sustainably. Typically, 60-70% of the energy content of the biomass is recovered in the producer gas. Using an average energy content for the biomass as 17 MJ/kg and a daily energy consumption of 2,300 MJ as in the case of the SOFC in table 6, the annual consumption of biomass is 49,400 kg. But considering that only 60% of the energy content is recovered in the producer gas the actual consumption is 82,300kg/year. With a demand of 82,300 kg/year the area required for sustainable production can be calculated once a few assumptions are made. The solar insolation of the area will be assumed to be 4.2 kWh/m2 per day and the main plants grown are C3 plants with a 10 year growing cycle. C3 plants are roughly 1.5% efficient in capturing the solar radiation so the incoming solar radiation converted to plant matter can be calculated to be 23kWh/ m2 over an entire year. Assuming these plants have a heating value of 17,000 kJ/kg, the area per kilogram required for plant production is 4.9 kg/m2-year. For the case of the SOFC in table 6, the required area for a 10 year plant harvested sustainably is 168,000 m2 or 16.8 hectares. For an economic analysis of this method of fuel production, the value of the land used for other purposes as well as labor costs should be analyzed. In many cases, this system could prove to be both economically viable and socially acceptable within the community.

Stephen Welty

Page 18

5/26/2002

Waste Heat and Exhaust Recovery The products of the reaction within the fuel cell to generate electricity are water and heat. In many cases the heat can be used for water heating or space heating. This application would be especially useful and practical in the on-site generation schemes discussed above where the heat does not have to be transported very far to get to its end users. Additionally, if the exhaust water is free of contaminants it can also be used for clean drinking water eliminating the need to boil water, which would in turn save a good deal of primary energy. In the case of Alkaline Fuel Cells, the exhaust water has been used for drinking water for astronauts during space missions. Depending on the technology used and the purity of the hydrogen supply, this may not be a feasible option. Waste Heat

1400

60

1200

50

1000

40

800 30 600 20

400

Recoverable Heat [kJ/s]

Water Exhaust [L/day]

Water L/day

10

200 0 0

500

1000

1500

0 2000

Energy [kWh/day]

Figure 6: Waste heat and water exhaust for different energy generation levels.

Figure 6 is a graph of the amount of exhaust water and recoverable waste heat for different levels of daily energy generation. For the case shown in table 9 for the SOFC the amount of water generated is only about ½ liter and the waste heat rate is 17 J/s assuming an 800 C temperature difference. Half of a liter is not enough drinking water for a family per day but at that heating rate 18L of water could be heated from 25C to 45C for bathing which could be sufficient for a family of two or three. In the case of the clustered community using a SOFC (table 6), the water generated is much more substantial. At the energy consumption rate of 640kWh/day, the amount of water generated from the fuel cell is about 367 liters. This amount of water would not be enough to supply the entire of community of 800 households with potable water but could be used to reduce the amount of water that must be boiled. In this case, the waste

Stephen Welty

Page 19

5/26/2002

heat could be recovered but it would be more difficult to transport it to the location of final use. There may be some other industrial demand in the community where the excess heat could be used. But as the electricity demand in the community grows, a gas turbine could be used in a combined cycle system to generate more electricity. This combination of a high temperature SOFC and a gas turbine has the highest potential for fuel to electricity efficiency of any electricity generation scheme currently available or projected to be available in the near future. However, this application is limited by the sizing of the gas turbines and is probably not available at such low steam generation levels as studied here. It is generally the case that once a community has a highly reliable and high quality electricity supply the electricity usage readily increases. As the electricity demand increases, so will the amount of water produced from the fuel cell per person, which would make recovery of the exhaust water a more practical proposition in the on-site generation case.

Conclusions Rural electrification has become a priority among a number of different governments and non-governmental organizations. As projects are identified it is important to note the time frame of the project as well as the potential sources of energy. If the time frame is long, it may be worth looking into cutting-edge technologies that may be available within the time frame of the project. With the increasing concern about environmental impacts and future possibilities of converting avoided carbon dioxide emissions into USD, renewable energies or more clean technologies could have a marked advantage economically as well. The most promising technologies for domestic electricity generation are SOFC and PEM. Both of these fuel cell technologies use solid electrolytes and are less susceptible to corrosion than their counterparts, which makes them more rugged and gives them longer lifetimes. The only commercially available fuel cell to date is a PAFC in the 200kW range. The main disadvantage of this technology is its 40,000 hour expected life. For a 20 year project the fuel cell would have to be replaced three times. SOFC technology has an advantage over PEM fuel cells in that it can tolerate a lesser purity of hydrogen and its high temperatures can be used to steam reform the methane into hydrogen. This translates into a lower cost and a higher efficiency for the SOFC technology since less energy is required to reform and purify the fuel. The higher temperatures of the SOFC also make waste heat recovery more of a possibility. However, PEM fuel cells have a higher energy density and a much more rapid response time, making them a favorite in the transportation sector. As PEM fuel cells are introduced into the transportation market, it will lead to a much higher volume of production than the SOFC, which in turn will reduce its cost. Fuel cells, although not yet readily available in the size ranges discussed in this paper, have had a great deal of development funding over the years. It is possible that within the

Stephen Welty

Page 20

5/26/2002

next few years these devices could be available in the size ranges discussed. If they become available and they meet the projected costs and lifetimes, they would have a clear advantage in a number of applications. Technology Diesel Genset PAFC SOFC PV/battery system PV/fuel cell system

Clustered Community 0.23$/kWh UCE 0.26$/kWh UCE 0.12$/kWh UCE 1.72$/kWh UCE 2.25$/kWh UCE

Scattered Community N/A N/A 1.34$/kWh UCE 2.86$/kWh UCE 2.79$/kWh UCE

Table # Table 4 Table 5 Table 6, 9 Table 7, 8 Table 10, 11

Table 12: Summary of electrification cases.

Table 12 is a summary of the cases investigated and shows that the SOFC would be the cheapest electricity alternative in the clustered community. In the case of on-site generation for the scattered community, the diesel generator was not analyzed because of noise and pollution problems. Instead, SOFC technology was compared to conventional PV technology and it was found that there is a clear economic advantage for the SOFC. Since the actual installed cost of the SOFC is unknown right now, an analysis of the effect of the installed cost on the UCE was carried out. It was found that even for installed costs of $5,500/kW, the UCE of the SOFC was below $0.50/kWh for the scattered community if yearly labor costs were not included (an assumption based on the fact that the owner could be trained to service the equipment). An additional advantage of fuel cells is their ability to be used as renewable energy conversion machines. If biogas or producer gas is used, the net carbon dioxide emissions from operation could be near zero. In the case of generating hydrogen from renewable energies, the case studied here of a photovoltaic system producing the gas proved to be uneconomical. However, using renewable energies to produce hydrogen should not be ruled out as it may prove to be a viable option if wind resources or hydro resources are available. Waste heat recovery and exhaust recovery in fuel cell systems provide for some interesting possibilities for enhancing energy efficiency in cogeneration and potable water that should not be overlooked. The possibilities for retrieving drinking water from the fuel cell exhaust will depend on the purity of the fuel input and the fuel cell technology used. Waste heat recovery for domestic use is more practical for the on-site generation schemes studied as the heat is close to the location of end use. However, even with the central generation station, the waste heat could be used in industry where the heat does not need to be transported very far.

Stephen Welty

Page 21

5/26/2002

Bibliography Brandon, Nigel and David Hart. An Introduction to fuel cell technology and economics. Imperial College of Science and Technology. Center for Energy Policy and Technology. July 1999. UK. Barra, Luciano. Hydrogen technology: status and perspectives. Class Handout. EG&G Services, Parson Inc., Science Applications International Corporation. Fuel Cell Handbook: Fifth Edition. US DOE. October 2000, Morgantown WV. Gagnon, Luc. Comparing Power Generation Options. Hydro-Quebec, directionEnvironment. April 2000. Websites: Smithsonian: National Museum of American History Behring Center. http://americanhistory.si.edu Brennstoffzellen in Jülich http://www.fuel-cell.de DOE Fossil Energy website. http://www.fe.doe.gov United Technologies Company http://www.utcfuelcells.com