Hydrogen from Renewable Energy Sources
Hydrogen from Renewable Energy Sources: Pathway to 10 Quads For Transportation Uses in 2030 to 2050 Task 3 Final Report October 2003
Prepared by:
Duane B. Myers, Gregory D. Ariff, Brian D. James, & Reed C. Kuhn
One Virginia Square 3601 Wilson Boulevard, Suite 650 Arlington, Virginia 22201 703/243-3383
Prepared for:
The Hydrogen Program Office Office of Power Technologies U.S Department of Energy Washington, D.C. Under Grant No. DE-FG01-99EE35099
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
FORWARD This work was funded by the Hydrogen Program Office of the U.S. Department of Energy under Grant No. DE-FG01-99EE35099 and represents the third and final task of three to be completed under this contract. The first task presented a broad overview of the costs for creating infrastructures to supply direct hydrogen, methanol, and gasoline to support fuel cell vehicles (FCV’s). A conclusion of the report resulting from the first task was that “the costs of maintaining the existing gasoline infrastructure per vehicle supported are up to two times more expensive than the estimated costs of maintaining either a methanol or a hydrogen fuel infrastructure”. The second task was to provide a detailed analysis of the cost of providing small-scale stationary hydrogen fueling appliances (HFA’s) for the on-site production and storage of hydrogen from natural gas to fuel hydrogen FCV’s. Four potential reforming systems were studied: 10-atmosphere steam methane reforming (SMR) with pressure-swing adsorption (PSA) as gas cleanup, 20-atm SMR with metal membrane gas cleanup, 10-atm autothermal reforming (ATR) with PSA gas cleanup, and 20-atm ATR with metal membrane gas cleanup. The third task, the subject of this report, was to identify a pathway for producing 10 quads of hydrogen for transportation uses per year from renewable sources in the years 2030 to 2050 (1 quad = 1015 Btu). The 10 quads of hydrogen is approximately the quantity that would be needed if all passenger vehicles were converted to hydrogen fuel cell vehicles. The scope of the analysis was limited to those categories of renewable sources that are potentially significant contributors on a 10 quad scale (wind, solar, geothermal, and biomass). We acknowledge the support of the Department of Energy, especially Sig Gronich, DOE Hydrogen Team Leader.
i
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
TABLE OF CONTENTS FORWARD ........................................................................................................................ i LIST OF FIGURES ......................................................................................................... iv EXECUTIVE SUMMARY .............................................................................................. 1 1 Basic Information on Renewable Energy Sources................................................. 6 1.1 What Does 10 Quads of Hydrogen Represent?.................................................. 6 1.1.1 Vehicle Energy Use ........................................................................................ 6 1.1.2 Natural Gas Pipeline Capacity ........................................................................ 6 1.1.3 Electricity Consumption and Grid Capacity ................................................... 7 1.2 Renewable Energy Sources for Hydrogen.......................................................... 7 1.2.1 Wind................................................................................................................ 8 1.2.2 Solar .............................................................................................................. 10 1.2.3 Geothermal.................................................................................................... 14 1.2.4 Biomass......................................................................................................... 15 2 Hydrogen Supply, Demand, and Economics ........................................................ 17 2.1 Renewable Energy Resources (Supply) ............................................................ 17 2.1.1 Competition for Renewable Energy Resources ............................................ 17 2.1.2 Electricity Grid Considerations..................................................................... 17 2.1.3 Wind.............................................................................................................. 18 2.1.4 Solar .............................................................................................................. 20 2.1.5 Geothermal.................................................................................................... 20 2.1.6 Biomass......................................................................................................... 21 2.1.7 National Totals.............................................................................................. 26 2.1.8 Other Sources of Renewable Energy for Hydrogen ..................................... 26 2.1.9 Other Issues................................................................................................... 27 2.2 Geographic Distribution of Hydrogen Demand............................................... 28 2.2.1 Per Capita Gasoline Consumption ................................................................ 28 2.2.2 State-by-State Population Projections........................................................... 28 2.3 Capital Costs........................................................................................................ 29 2.3.1 Wind.............................................................................................................. 30 2.3.2 Solar .............................................................................................................. 30 2.3.3 Geothermal.................................................................................................... 31 2.3.4 Biomass to Hydrogen.................................................................................... 31 2.3.5 Biomass-to-Electric....................................................................................... 32 2.3.6 Landfill Gas .................................................................................................. 33 2.3.7 Hydrogen Pipelines....................................................................................... 33 2.3.8 Interstate Electricity Transmission ............................................................... 34 2.3.9 Electrolyzers ................................................................................................. 34 2.3.10 H2 Filling Station (Compression, Storage, and Dispensing)......................... 35 2.4 Hydrogen and Electricity Cost Models ............................................................. 35 2.4.1 Renewable Electricity Generation ................................................................ 35
ii
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.2 Electrolysis.................................................................................................... 38 2.4.3 Biomass Gasification and Reforming to Hydrogen ...................................... 39 2.4.4 Electricity Transmission Lines and Hydrogen Pipelines .............................. 40 2.4.5 Hydrogen Filling Stations (Compression, Storage, and Dispensing) ........... 41 2.4.6 Cost Summary............................................................................................... 41 3 Proposed Hydrogen Generation Network ............................................................ 43 3.1
Development of Hydrogen Distribution and Cost Model................................ 43
3.2 Results .................................................................................................................. 46 3.2.1 National Statistics ......................................................................................... 46 3.2.2 Selected State Statistics................................................................................. 51 3.3 Sensitivity Study for Important Variables ....................................................... 55 3.3.1 Hydrogen demand......................................................................................... 55 3.3.2 Low-cost solar............................................................................................... 56 3.3.3 Low-cost electrolyzed hydrogen................................................................... 57 3.3.4 Low-cost electricity transmission ................................................................. 58 3.4 Alternative Scenarios.......................................................................................... 59 3.4.1 Electrolysis at centralized facilities .............................................................. 59 3.4.2 Electrolysis at location of resource ............................................................... 60 3.4.3 Electricity from Renewables, Hydrogen from Natural Gas.......................... 61 3.4.4 Nuclear Hydrogen......................................................................................... 63 3.5 Study Limitations................................................................................................ 64 4 Conclusions.............................................................................................................. 66 APPENDIX A: HIGHER BIOMASS AVAILABILITY ESTIMATES..................... 67 APPENDIX B: MODEL INPUT - RENEWABLE ENERGY RESOURCES BY STATE IN 2040 ............................................................................................................... 70 APPENDIX C: MODEL OUTPUT - STATE BUYING HISTORIES FOR BASELINE ...................................................................................................................... 71
iii
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
LIST OF FIGURES Figure 1-1. Classes of Wind Power Density...................................................................... 9 Figure 1-2. Wind Class Distribution in the United States. (Source: Wind Energy Resource Atlas of the United States, available at http://rredc.nrel.gov/wind/pubs/atlas/maps/chap2/2-01m.html)............................... 10 Figure 1-3. Average solar radiation during the month of June in the U.S. (units are kWh/day-m2)............................................................................................................. 11 Figure 1-4. Solar radiation striking a fixed flat plate tilted toward the south at the latitude angle for selected U.S. cities..................................................................................... 11 Figure 1-5. PV solar-electric examples............................................................................ 13 Figure 2-1. Livestock manure generation by animal type. Wet manure is approximately 14.5% dry matter, and hydrogen yield is 7% by weight on a dry basis.................... 25 Figure 2-2. Summary of hydrogen potential from renewable sources............................. 26 Figure 2-3. Correlation between gasoline usage and population by state (1999 data for gasoline usage, 2000 data for population). ............................................................... 29 Figure 2-4. The cost of photovoltaic solar systems (including solar module and balance of system components) by year of installation. The individual data points were taken from several reports from actual PV solar installations., , .............................. 30 Figure 2-5. Biomass gasification capital costs for the atmospheric pressure Battelle steam gasification process, adapted from Reference 82. The costs have been adjusted to 2001 dollars. ........................................................................................... 32 Figure 2-6. Biomass-to-electricity plant capital costs for a 122 MWe biomass gasification/combined cycle system. Adapted from Reference 83, adjusted to 2001 dollars........................................................................................................................ 32 Figure 2-7. Cost of high-voltage electricity transmission lines as a function of capacity. ................................................................................................................................... 34 Figure 2-8. Discounted cash flow parameters for renewable electricity generation (solar, wind, geothermal). .................................................................................................... 36 Figure 2-9. Discounted cash flow parameters for biomass gasifier electricity generation. ................................................................................................................................... 36 Figure 2-10. Discounted cash flow parameters for landfill gas electricity generation. .... 37 Figure 2-11. Discounted cash flow parameters for electrolyzer distributed hydrogen generation.................................................................................................................. 38 Figure 2-12. Discounted cash flow parameters for biomass gasification and reforming to hydrogen in large-scale centralized plants................................................................ 39 Figure 2-13. Discounted cash flow parameters for interstate electricity transmission and hydrogen pipelines. ................................................................................................... 40 Figure 2-14. Discounted cash flow parameters for hydrogen filling station compression, storage, and dispensing components......................................................................... 41 Figure 2-15. Cost factors used to calculate the cost of hydrogen in Section 3. ............... 42 Figure 3-1. Hydrogen costs from the various production methods. ................................ 45 Figure 3-2. Cost breakdown for hydrogen from (a) renewable electricity and (b) reformed biomass...................................................................................................... 45 Figure 3-3. Model resource availability and usage (in quads of hydrogen per year). ..... 47 Figure 3-4. Summary of hydrogen cost (w/ 500 mile transmission), availability, and usage for each resource predicted by the model. ...................................................... 47
iv
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Figure 3-5. National hydrogen cost distribution. ............................................................. 48 Figure 3-6. Evolution of national average hydrogen cost throughout purchase process. Costs reflect hydrogen purchased in a given round and not the effect of hydrogen from previous rounds. ............................................................................................... 49 Figure 3-7. Average cost per kilogram of hydrogen for each state. ................................ 49 Figure 3-8. Distribution of transmission distances for hydrogen by pipeline.................. 50 Figure 3-9. Distribution of electricity transmission distances for electrolyzed hydrogen. ................................................................................................................................... 50 Figure 3-10. Hydrogen purchases (in order of preference) for Texas, North Dakota, California, and New Jersey. ...................................................................................... 52 Figure 3-11. Summary of source states and resources for hydrogen consumed by Texas, North Dakota, California, and New Jersey. .............................................................. 53 Figure 3-12. Texas (a) hydrogen cost distribution, (b) resource usage. .......................... 53 Figure 3-13. North Dakota (a) hydrogen cost distribution, (b) resource usage. .............. 54 Figure 3-14. California (a) hydrogen cost distribution, (b) resource usage..................... 54 Figure 3-15. New Jersey (a) hydrogen cost distribution, (b) resource usage. ................. 55 Figure 3-16. Cost of hydrogen from renewables as a function of national demand for 1, 3, 5, 10, and 15 quads/year. ...................................................................................... 56 Figure 3-17. Resource usage pies for scenarios in which solar electricity costs (a) equal wind Class 4 and (b) equal wind Class 6. ................................................................. 57 Figure 3-18. Results from scenarios with reduced costs for solar hydrogen. .................. 57 Figure 3-19. Resource uses and potential for three electrolyzed hydrogen costs........... 58 Figure 3-20. National resource usage and potential for three electricity transmission costs........................................................................................................................... 58 Figure 3-21. Summed Class 5 and 6 wind resource usage for reduced electricity transmission costs. .................................................................................................... 59 Figure 3-22. At-resource versus at-station electrolysis cost comparison (in $/kWh) for hydrogen production from Class 6 Wind.................................................................. 61 Figure 3-23. Costs of renewably generated electricity. ................................................... 62 Figure 3-24. Estimated cost and CO2 emissions of alternative renewable electricity and hydrogen from natural gas compared to the model baseline. ................................... 63 Figure 3-25. Hydrogen costs from renewable and non-renewable sources, including transmission, compression, storage, and dispensing costs........................................ 65
v
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
EXECUTIVE SUMMARY Introduction This report is the final in a series of studies by Directed Technologies, Inc. of the cost and infrastructure requirements to supply hydrogen for fuel cell vehicles (FCVs). The previous studies have concentrated on the early development of the hydrogen infrastructure to supply a limited number of FCVs. In this study, the focus shifts to future years in which a significant fraction of passenger cars and trucks are FCVs. The goal of the current analysis was to develop a technically feasible pathway to supply 10 quads1 per year of hydrogen fuel from renewable energy sources for transportation uses in the years 2030 to 2050. The midpoint year 2040 was used for calculations of resource availability and demand that may change over the 20 year period. To put 10 quads of hydrogen energy in perspective, if the passenger vehicles currently on the road were converted to FCVs, 10 quads of hydrogen would be sufficient to fuel all of those vehicles.2 Approach and Results The first task was to determine the total potential hydrogen generation from renewable resources in the U.S. in 2030-2050 to ensure that the production of 10 quads of hydrogen was possible. All renewable resources were considered; however, only the resources that could make a contribution of at least 0.1 quads per year were included: biomass, solar, wind, and geothermal. Hydrogen from the biomass resources was produced through gasification and steam reforming at centralized facilities with pipeline delivery to the service stations. Hydrogen from the electricity-generating resources was produced by water electrolysis at the service stations.3,4 Hydrogen service stations were assumed to have onsite compression, storage, and dispensing for 5000psi delivery to the vehicle. The potential hydrogen generation from each of the four resources were determined on a state-by-state basis. The U.S. totals for hydrogen potential in 2040 are listed in Figure 1. The hydrogen potential includes the effect of conversion efficiencies (gasifier or electrolyzer) and line losses in transmission. The hydrogen demand for a given state in the year 2040 was estimated by assuming that the per capita demand for hydrogen would be proportional to the per capita gasoline usage in that state.5 To calculate the gasoline need for a state in 2040, we multiplied the predicted population by the current per capita gasoline usage, assuming that the relative gasoline use per capita remains the same for each state in the future. The 10 quads of hydrogen was then divided among the states in the same proportion as the projected gasoline consumption. 1
One quad = 1015 Btu. In this report, the lower heating value (LHV) is used when referring to the energy content of a fuel. 2 Based on an average 2.2X efficiency gain for FCVs over conventional internal combustion engines. 3 Alternate hydrogen production methods were considered, but the lowest cost routes to hydrogen were gasification/reforming from biomass and electrolysis from wind, solar, and geothermal electricity. 4 As an exception to these methods, landfill gas was assumed to be delivered to the service stations via natural gas pipelines and reformed onsite. 5 The per capita gasoline consumption varies considerably from state to state.
1
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
The cost for hydrogen from each resource was calculated by applying discounted cash flow (DCF) analysis for each step along the hydrogen pathway from production to dispensing into vehicles. The capital costs for equipment were taken from the literature or from projections for future costs for renewable technologies. The factors used to calculate the cost of hydrogen are listed in Figure 2. Hydrogen Energy Potential in 2040 quads/year Biomass
2.76
Wind
22.9
Solar
5.9
Geothermal
0.4
U.S. Total
31.9
Figure 1. Hydrogen potential from the major renewable resources in 2040.
Resource Wind Class 4 Wind Class 5 Wind Class 6 PV solar Geothermal Biomass Gasification Electricity transmission (high-voltage) Interstate H2 pipeline Local H2 pipeline Electrolyzer Compression, storage, and dispensing
Capital Costa $648/kWp $648/kWp $648/kWp $1,000/kWp $1,500/kWp $71,091/m.t. H2 per day ($102,000 + $500/MW) per mile $2.79/(mile-106 Btu/day) N/Ac
Capacity Factor 38.3% 41.4% 48.7% 25% 95%
Operating Costsb $0.019/kWh $0.018/kWh $0.016/kWh $0.032/kWh $0.012/kWh
85%
$0.83-$1.26/kg H2d
70%
$0.0016/kWh
60%
$0.064/106 Btu
N/Ac
$2.99/106 Btu
$300/kWth
69%
$0.87/kg H2
$513/(kg H2/day)
69%
$0.27/kg H2
a
Capital costs are in 2001 dollars. Includes operation & maintenance, utilities, and overhead expenses. c Local H2 pipeline cost per unit energy was assumed to be the same as the current cost for local natural gas pipeline. d Biomass operating cost includes feedstock costs ranging from $22/metric ton (for manure and MSW) to $44/metric ton (for dedicated energy crops). b
Figure 2. Factors used to calculate the cost of hydrogen from renewable resources.
6
The H2 potential from biomass listed here accounts for competition for biomass resources for alternate uses, with a maximum of 4.3 quads. See Section 2.1.6 and Appendix A of the full report for details.
2
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
The resulting costs of hydrogen at the dispensing site from the various pathways are shown in Figure 3. The hydrogen costs listed on Figure 3 include the cost for 500 miles of transmission for comparison purposes, although the distribution system described below accounts for actual transmission miles.
7
a
Electricity Electrolysis Comp.,Stor.,Disp. Transmission (per 500 miles)
3.5 3
5
H2 cost, $/kg
H2 cost, $/kg
6
4
2 1.5
2
1
1
0.5
Wind Class 4
Wind Class 5
Wind Class 6
PV Solar
GeoThermal
Feedstock Gas. & Ref. Local H2 pipeline Local NG pipeline Comp.,Stor.,Disp. Pipeline (per 500 miles)
2.5
3
0
b
0
Dedicated Agricultural Wood Energy Residues Waste Crops Residues
MSW
Livestock Manure
Landfill Gas
Figure 3. Cost of hydrogen from renewable resources, including 500 miles transmission from source to vehicle dispensing for (a) electrolysis and (b) reformation production methods.
A simulation was created to generate a hydrogen production scenario for the continental U.S. based on resource availability and cost. The estimated costs of hydrogen are based on generation and transportation costs, with interstate distances taken into consideration. The model strives to simulate an unregulated hydrogen market in which consumers in each state buy hydrogen generated from the least expensive available resource from any other state. The algorithm takes as input the resource availability, base costs, and needs in each state and outputs the resulting distribution of hydrogen resource usage and prices. While the simulation does not deliver the true cost-optimized renewable hydrogen market, it does model a realistic low-cost hydrogen supply infrastructure. In the simulation, hydrogen is purchased in small units (0.0001 quads), with each state purchasing the cheapest available hydrogen in each buying round until all of that state’s hydrogen needs are met or the resources are consumed. At the beginning of each round, a cost for hydrogen generated using each resource from each state to each state is calculated. If a particular resource in a state is non-existent or has been consumed, that resource is unavailable for purchase. The simulation generates a round-by-round history of the hydrogen purchase choices for all states, providing potential resource preferences, locations, and transmission. In order to evaluate the simulation results, we can look at the statistics for hydrogen production on a national basis as well as examining the results for individual states. Figure 4 shows the simulation results for the annual resource availability and consumption. Consumption of each resource depends both on its availability and on the
3
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
cost of hydrogen from that resource relative to all other resources. While biomass and geothermal resources represent a small hydrogen potential relative to solar and wind energy, these resources make significant contributions to the national hydrogen demand due to their relatively low cost compared to other renewable resources. Solar energy, however, makes no contribution since it is considerably more costly than the abundant Class 4 wind energy with which it must compete. Of particular note is that 3.34 quads of the 4.8 quads of potential hydrogen from wind classes 5 and 6 goes untapped. Much of this unused wind potential is in the Midwest and Rocky Mountains, where the high transmission cost to the population centers on the coasts prohibits higher-class wind from being cost competitive with Class 4 wind resources at shorter transmission distances. Resource
Potential (quads/yr)
Usage (quads/yr)
Wind Class 4
18.1
5.3
Wind Class 5
3.1
0.48
Wind Class 6
1.7
0.98
Photovoltaic Solar
5.9
0.0
Geothermal
0.43
0.43
Dedicated Energy
0.57
0.57
Agricultural Residue
0.81
0.81
Wood Waste
0.73
0.73
Municipal Solid Waste
0.18
0.18
Livestock Manure
0.29
0.29
Landfill Gas
0.13
0.13
Figure 4. Summary of hydrogen availability and usage for each resource predicted by the model.
The resulting average cost of hydrogen in each state is listed in Figure 5. Generally, the least expensive hydrogen initially available to all states is from in-state biomass resources, followed by the biomass in neighboring states. Nearly half of all hydrogen from biomass is consumed in-state, with the amount transported out of state (via pipeline) decreasing with increasing inter-state distance. The average inter-state pipeline distance is 259 miles. As states consume their surrounding biomass resources, they are forced to purchase electricity for hydrogen generation, which may be transmitted over longer distances. Only 24% of renewable electricity is used in-state for hydrogen production, and no obvious trend exists for quantity of hydrogen transmitted as a function of transmission distance. The mean distance for renewable electricity transmission is 540 miles. It is interesting to note that no resource is transmitted (via electrical lines or pipelines) more than 1,500 miles, indicating that coastal states reach no further than the middle of the country to meet their hydrogen demand.
4
Hydrogen from Renewable Energy Sources
AL AR AZ CA CO CT DE FL GA IA ID IL IN KS KY LA
$3.92 $3.72 $3.49 $4.09 $3.51 $3.25 $2.60 $4.72 $4.49 $3.11 $2.73 $4.16 $4.01 $3.23 $3.71 $3.63
Directed Technologies Inc.
MA MD ME MI MN MO MS MT NC ND NE NH NJ NM NV NY
$4.17 $4.11 $2.73 $4.17 $3.60 $3.84 $3.48 $2.60 $4.50 $2.54 $2.62 $2.72 $4.45 $3.08 $3.03 $4.52
OH OK OR PA RI SC SD TN TX UT VA VT WA WI WV WY
$4.30 $3.36 $3.33 $4.39 $2.62 $3.98 $2.56 $4.11 $4.03 $3.13 $4.34 $2.57 $3.68 $3.70 $2.64 $2.59
Figure 5. Average cost per kilogram of hydrogen (delivered to vehicle at 5000psi) for the 48 contiguous states. The U.S. average cost is $3.98 per kg of H2.
Conclusions The annual generation of 10 quads of hydrogen in the years 2030-2050 from renewable sources for transportation uses in the U.S. is technically achievable and, according to our model, leads to a national average hydrogen cost of $3.98/kg ($33.24/GJ, LHV basis) delivered to the vehicle. Wind and biomass are the most significant resources (on an energy supplied basis) for hydrogen production, with geothermal playing a small role due to its limited potential. Hydrogen from renewable electricity is expensive compared to that from the reformation of biomass for three important reasons. First, most renewable electricity, especially wind and solar, is expensive compared to electricity from fossil fuels due primarily to high capital costs for the wind and solar installations with relatively low capacity factors (all 150°C, geothermal source to be economically viable. However, there is a significant resource in the U.S. of low-quality geothermal energy that may be economical for direct uses such as space heating, agriculture, and aquaculture. Such application could displace the use of fossil fuels used for heating (direct or through electricity generation), freeing those fuels for hydrogen production in processes such as steam methane reforming. Since this report focuses on hydrogen production from renewable resources, only geothermal sources appropriate for electricity production will be considered. There are three methods of geothermal electricity generation, listed in order of generally decreasing reservoir temperature and correspondingly lower theoretical efficiencies: •
32
Direct steam: Naturally-occurring steam from a geothermal well is used to drive a turbine. Geothermal fluids typically contain impurities, including salts and sulfur compounds, that must be removed or treated before disposing of the condensate.
“Generating Power with the World’s Tallest Tower.” Business Week, 9 September 2002, p. 69.
14
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
•
Flash steam: Similar to direct steam except that the geothermal fluid is high temperature water or a mixture of water and steam that must be “flashed” (exposed to a pressure drop) to separate the usable steam from the water.
•
Binary: The hot geothermal resource (steam, hot water, or two-phase mixture) is used to vaporize a working fluid that has a boiling point lower than water. Isobutane, pentane, and hydrofluorocarbon (HFC) refrigerant fluids are common working fluids in binary plants.33
The direct steam process is the most economical method of the three since steam from the geothermal reservoir can be used with little pretreatment. Flash steam and binary are less economical since more processing steps, and consequently additional equipment, are required for heat recovery from the reservoir. Commercial installations of geothermal electricity plants are generally small compared to utility power plants, with the exception of The Geysers plant constructed in California in the 1960’s. The Geysers nominal capacity is 2,850 MW, or the equivalent of four large coal- or natural gas-fired power plants. The actual power output from The Geysers has declined over the years due to depletion of the steam reservoir. The trend in geothermal electricity generation is toward small distributed plants and away from power-plant level production. 1.2.4 Biomass “Biomass” is generally considered to be any material that is derived from plant or animal matter so that the use of the material for energy generation, which by its nature causes emissions of carbon dioxide, can be offset by the growth of new plant or animal matter to reabsorb a quantity of carbon dioxide equal to that emitted. For the purposes of this study, we have considered the following categories of materials to be biomass:
•
•
•
Energy crops: A plant that is raised specifically to use for energy generation is considered an energy crop, and theoretically any plant could be used for energy or hydrogen generation. The plants most often associated with energy production are switchgrass, hybrid poplar, and willow. Agricultural residues: Stalks that remain in the field after harvest are classified as agricultural residues. Sustainable farming practices require that some crop residue be left in the field to supply organic material to the soil necessary for the next planting season. There may be some residues from food processing (e.g., peanut shells) that could be converted to hydrogen, but these quantities are likely to be small in comparison to harvest residues. Wood waste: Material left in the forest after timber harvesting and the discard such as sawdust from processing are considered wood waste. Much of the waste in saw mills and paper making facilities is already burned to produce steam or
33
Rafferty, Kevin. Geothermal Power Generation: A Primer on Low-Temperature, Small-Scale Applications. Report from the Geo-Heat Center, January 2000.
15
Hydrogen from Renewable Energy Sources
•
•
Directed Technologies Inc.
process heat for the facility, so the availability of this biomass will depend on the ability to find substitute sources of steam and process heat. Municipal solid waste (MSW) and landfill gas: It can be argued that MSW and landfill gas are not renewable in the same way that solar or wind energy is renewable, however, we have included both in this analysis since in most cases the waste would be unused for any other purpose. Livestock manure: High population-density livestock facilities such as dairy farms, hog farms, and chicken houses handle and process large quantities of manure. There is increasing pressure on farmers in some locations to reduce odors from and discharge of livestock wastes to waterways. In many locations the manure must be collected and processed before disposal or land application as fertilizer, so the infrastructure may already be in place to deliver the waste to a central hydrogen production facility.
In the case of biomass resources, three potential pathways to hydrogen exist: • •
•
The first pathway is biomass gasification combined with a gas turbine or with a conventional steam-cycle to generate electricity, in which case the electricity is treated in the same way as solar, wind, or geothermal energy. The second pathway is gasification and steam reformation, generating hydrogen directly from the biomass. Distribution of hydrogen then generally requires a hydrogen pipeline network, possibly supplemented by liquid hydrogen transportation by truck for locations not served by the hydrogen pipeline. The third pathway is conversion of the biomass to liquid hydrocarbons, or biocrude, near the feedstock source, with the biocrude transported by truck or pipeline to biocrude steam reformers at the hydrogen dispensing sites.
The one exception to these cases is reformation of landfill gas. Since landfill gas is predominantly methane and carbon dioxide, the methane can be purified and routed directly into the natural gas pipeline. An equivalent volume of natural gas can be removed from the pipeline at the refueling site, where it can be steam reformed to produce hydrogen. The biomass-to-H2 pathway chosen for a particular feedstock and location will depend on the economics of the distribution system. The pathway for a particular location will be dependent on the availability and cost of distribution infrastructure, and these issues are addressed in detail in Sections 2 and 3.
16
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2 Hydrogen Supply, Demand, and Economics 2.1 Renewable Energy Resources (Supply) The first task in developing a 10 quad/year hydrogen-from-renewables system is to determine that at least 10 quads of H2 are available. This section provides the estimates for total potential hydrogen from each of the renewable energy sources discussed in Section 1.2. Generally, the literature on renewable energy list sources by state, and we aggregated the state numbers to develop the national totals. We have chosen 2040 to be the benchmark year for calculating renewable resource availability (midpoint of the 2030-2050 period), since some resources increase or decrease every year. The potentials of the major renewable resources in this analysis do not depend on the year, as the availability of the resources is only marginally influenced by the state of technology. The quantity of feedstocks such as municipal solid waste, however, are directly proportional to the population and will change every year.
The estimates presented in this section are single values for each of the resources, although there is significant uncertainty in projections for 30 to 50 years in the future. In Section 3 we examine the sensitivity of our predictions to the important assumptions about cost and availability. 2.1.1 Competition for Renewable Energy Resources The use of renewable resources for applications like electricity generation may infringe on the potential resource available for hydrogen generation. Federal and state governments are actively developing or considering requirements for renewable energy portfolios in electricity generation. The U.S. Senate recently approved a provision in the energy bill that would require investor owned utilities to produce a minimum of 10% of their electricity from renewables by 2020.34 With a projected electricity generation of 16.8 quads,35 of which 79.4% would be supplied by investor owned utilities,36 the potential need for renewable electricity to support the regulation is 1.3 quads in 2020. The quantity of electricity needed to meet the 10% requirement may be as high as 2 –3 quads in 2030-2050 with modest growth in electricity usage and an increase in renewables contribution to the electricity supply. 37 2.1.2 Electricity Grid Considerations The major sources of renewable energy, with the possible exception of biomass, directly generate electricity, with the electricity converted to hydrogen by electrolysis of water. Since electricity is prohibitively expensive to store in quantities on the scale of this analysis, the electricity must be used when it is generated. In this study we trace the “electrons” from the source through the transmission systems to hydrogen only to calculate the cost of hydrogen from a particular renewable electricity source.
34
U.S. Senate, “Energy Policy Act of 2002,” April 25, 2002. Based on H.R. 4. Department of Energy, Energy Information Administration, Annual Energy Outlook 2002. 36 Department of Energy, Energy Information Administration, Annual Energy Review 2000. August 2001. 37 The U.S. DOE in the Annual Energy Outlook 2003 projects renewable electricity generation (excluding hydro) to be 131 billion kWh (~0.45 quads) in 2025 in the absence of new federal mandates. 35
17
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
To simplify our estimates, we have ignored two important realities of renewable electricity production: 1) production intermittency for wind and solar resources as it affects grid capacity and electricity demand; 2) peak versus off-peak electricity pricing. Ignoring the first issue means that hydrogen prices reflect the average cost of electricity production at the renewable resource site and the cost of transmitting electricity for electrolysis to the service station, but not the cost of extra grid capacity to account for the sporadic electricity production. Ignoring the second means that while the actual electricity produced from the resources can be used by any electricity-consumer on the grid, the cost of renewable electricity is paid for by hydrogen consumers alone. In reality, hydrogen production will be timed to coincide with periods of low electricity prices, assuming that there is enough buffer in the hydrogen storage system to outlast the periods of high electricity prices. For example, solar electricity is generated only during daylight hours and could be sold to electricity users at a favorable price during peak-use hours, but hydrogen production may be scheduled at night when the cost of electricity from the grid is lower. We have also assumed that the electricity grid is stable, that is, it provides a consistent voltage and load that meets the demand, no matter what the mix of renewable sources of hydrogen. Wind and solar are both intermittent resources that can not be reliably predicted on a continuous basis, so there is a maximum contribution from renewables that can be compensated for by the non-renewable electricity production methods. Denmark, a country with huge offshore wind electricity installations, has experienced grid instability at about 20% contribution from wind.38 The efficiency of electrolyzers is defined as the energy content of the hydrogen divided by the electrical energy input. The projected efficiency based on the hydrogen lower heating value (LHV) of future electrolysis processes is anticipated to be 76%.39 At this projected efficiency, 1.32 units of electricity are required to generate each unit of hydrogen energy in the electrolyzer. Including typical electricity transmission line losses of 6% system-wide,40 1.4 units of electricity must be generated at the source to produce one unit of hydrogen energy at the electrolyzer. 2.1.3 Wind Our estimate for hydrogen recoverable from wind energy is based on the following assumptions:
•
All U.S. land of Classes 4, 5, and 6 wind in the contiguous 48 states and within 10 miles of 100 kV or greater transmission lines is utilized, with the following exclusions: parks, urban areas, water, 50% of forest land, 30% of agricultural land, and 10% of range land.41 This assumes all qualifying land is put to use,
38
Insert the Denmark reference. Ogden, J.M. and J. Nitsch. “Solar Hydrogen,” in Renewable Energy: Sources for Fuels and Electricity. Johansson, Kelly, Reddy, and Williams, eds. Island Press, Washington, D.C., 1993. 40 U.S. Department of Energy, Energy Information Administration, Annual Energy Review 2000. 41 EPRI and U.S. Department of Energy. Renewable Energy Technology Characterizations, TR-109496, December 1997. 39
18
Hydrogen from Renewable Energy Sources
• • • •
Directed Technologies Inc.
although a maximum of only 5% of the land is actually covered by the wind turbines and associated equipment. The remaining 95% of the land area can still be dedicated to other uses, for example, farming, livestock grazing, or even solar electricity generation. The contribution of each wind class to an individual state’s total wind resource was determined from the wind resource map (Figure 1-2). Capacity factors42 are 38.3% for Class 4, 41.4% for Class 5 and 48.7% for Class 6 based on projections for improved technology.43 Availability factor44 is 98%. Turbines are assumed to be large-scale (1 MW each), with a 50-meter minimum hub height, the standard for larger turbines.
After accounting for an electrolyzer efficiency and transmission losses, the quantity of hydrogen that could be produced from wind energy is • • • •
Class 4: Class 5: Class 6: Total:
18.1 quads 3.1 quads 1.7 quads 22.9 quads
Electricity generation from wind must grow at a 15-16% compound annual rate to increase from the 2000 level of less than 0.1 quad45 to the 2040 level of 22.9 quads. Sustaining such a high growth rate in wind power installations may be difficult, since the highest quality, and consequently lowest cost, wind sites will be developed first, leaving the lower quality sites undeveloped. However, since we are developing a pathway to only 10 quads of H2, a maximum of 45% of the wind resources will be tapped even if no other renewable energy sources are included. There are substantial sources of Classes 5-7 wind off the Oregon and Massachusetts coasts, which could yield powerfully efficient energy pending public aesthetic acceptance, advancements in offshore turbine technology, and ease of access. The offshore wind resources have not been included in the totals listed above due to the low probability of significant offshore wind development by 2030. The outcome of current and near-term offshore wind projects throughout Northern Europe will likely determine if the U.S. will follow suit and utilize these resources. The projections used to estimate the wind potential did not assume any extension of the electricity grid to reach additional remote locations. It is likely that by 2030-2050 there will be more Classes 5 and 6 land areas that meet the 10 miles-to-transmission line qualification. Although the total Classes 5 and 6 potential would not reach the levels of
42
“Capacity factor” is the ratio of the average rate of power production to the theoretical capacity. Ref. 41. 44 “Availability factor” is the fraction of time the generator is operational, considering maintenance downtime. 45 Energy Information Administration. Annual Energy Review 2000. 43
19
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Class 4, the lower costs of Classes 5 and 6 electricity compared to Class 4 (see Section 2.4) may influence the decision to use wind instead of other resources. Furthermore, the capacity factors for these wind classes are expected to improve with better turbine engineering and advances in rotor materials and design. Capacity factors in this study were taken from 2001 projections for 2030, although there have already been demonstrations of 48-52% capacity factors.46 2.1.4 Solar The only practical restrictions on the availability of solar resources are the cost (to be considered in Sections 2 and 3) and the land area required. The favorable locations for using PV solar electricity to produce hydrogen are in the desert Southwest (e.g., Arizona, New Mexico, southern California) since other southern locations (e.g., Florida, Texas) have more cloud cover and an insolation approximately 10-15% lower than in the desert region. On a national level, only the most efficient solar resources will be exploited for utility-scale installations that are required to make an impact on the 10-quad scale. Rooftop solar panels on commercial buildings and private residences are more likely to generate electricity for local consumption (offsetting consumption of electricity from power plants) than to provide electricity to the grid for hydrogen production.
The Bureau of Land Management (BLM) controls 262 million acres in the western U.S., with approximately 22.3 million acres (8.5%) of that land in prime solar areas that have average annual insolation > 6 kWhr/m2-day. The land area required to generate one quad of hydrogen in these regions is projected to be about 378,000 acres.47 If 10% of the BLM land is used for solar electricity generation, the quantity of hydrogen that could be generated is 5.9 quads. It is uncertain whether or not nearly 1% of BLM land will be developed for solar electricity generation, however, this projection provides a reasonable range for solar potential. As we demonstrate in Section 3, unless the cost of solar electricity can be reduced below the level projected for 2040, there will be few, if any, utility-scale solar electricity generation installations and land availability will not be an issue.48 2.1.5 Geothermal The best estimate of the geothermal potential for electricity generation in the U.S. with optimistic assumptions about improvements in drilling and reservoir management
46
“The Potential of Wind, Solar, Geothermal for the West”, Dr. A. Leitner, RDI Consulting, NEMS/Annual Energy Outlook 2002 Conference. 47 Assumes a 25% solar-to-electric conversion efficiency (an optimistic projection for future technology), 50% land coverage for solar modules and balance of system equipment, 76% electrolyzer efficiency (LHV basis), and 6% electrical transmission line losses. 48 In 2030-2050, solar electricity may be developed enough to play a significant role in distributed power generation or summer peaking, since the cost of solar electricity may be competitive with peak prices in some markets. However, solar electricity is unlikely to be cost-competitive with wind when there is a choice between the two.
20
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
technologies is 18,800 MW.49 We assume that the geothermal electric capacity has been realized by 2030 and remains essentially flat through 2050. The hydrogen yield from electrolysis of 18,800 MWe operated 24 hours/day, 365 days/year is 0.43 quads. The lack of cooling water in desert locations, especially in Nevada and Southern California where geothermal resources are most prevalent, may limit the number of locations that can be exploited for geothermal electricity generation since cooling tower water is a much preferred heat rejection medium than is air. The cooling water can theoretically be cooled to the ambient wet bulb temperature, which is often substantially below the air temperature in arid climates. The theoretical maximum energy efficiency calculated using Equation 1-6 increases approximately 2.4% for every 10°C reduction in cooling fluid temperature.50 2.1.6 Biomass In the years 2030-2050 there are likely to be many competing uses for biomass resources and land, including electricity generation by co-firing biomass with fossil fuels, conversion to liquid fuels such as ethanol and biodiesel, and increased demand for food as the population increases. Accurately projecting the demand for any biomass use 30 to 50 years in advance is difficult, so a logical and defensible allocation of biomass resources must be estimated. The biomass resource allocation method used in the subsections to follow is based on limiting availability by capping the biomass feedstock costs. Capping the cost of each biomass resource for the production of H2 for transportation uses frees the remaining biomass for other end uses that are not a concern of this analysis.
To calculate the quantity of hydrogen that could be generated from biomass, we have assumed that the biomass is gasified and steam reformed to produce pure hydrogen. This pathway produces the highest hydrogen yield per unit of biomass, so any other process that produces hydrogen from biomass will have a lower hydrogen potential. Based on simulations of the gasification process, approximately 70 kg H2 (net) can be produced per dry metric ton of biomass. This yield is consistent with literature values from commercial gasification processes51 and is constrained primarily by thermodynamic limitations that are unlikely to be overcome by improvements in technology in future years.52
49
Gawell, K, et al. Preliminary Report: Geothermal Energy, the Potential for Clean Power from the Earth. Geothermal Energy Association, April 1999. 50 Calculated for a 150°C reservoir temperature with 35°C, 25°C and 15°C cooling fluid temperatures. The resulting efficiencies were 27.2%, 29.6%, and 31.9%, respectively. 51 Craig, K.R. and M.K. Mann. Cost and Performance Analysis of Biomass-Based Integrated Gasification Combined-Cycle (BIGCC) Power Systems, National Renewable Energy Lab Report, October 1996. 52 There is a possibility that integrated gasifier/separation processes-for example, membrane reactors-that remove H2 from the reaction zone to force the equilibrium toward higher H2 production will be developed by 2030. However, we do not anticipate that membrane reactors will contribute a significant fraction of the for the large-scale installations required for widespread H2 production.
21
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.1.6.1 Dedicated Energy Crops
De La Torre Ugarte, et al. conducted a modeling study to determine the potential for conversion of existing agricultural land to bioenergy cropland by the year 2008.53 The model calculates the effects of competition between energy crops and existing crops, livestock grazing, and Conservation Resource Program (CRP) land. The land area that would be devoted to energy crops in the model scenario is 41.87 million acres with a total energy crop of 188.1 million dry tons for a yield of 4.5 dry tons per acre at a price of $50/dry ton delivered. In this case, the only energy crop expected to be economically competitive is switchgrass, with all production east of the Rocky Mountains. The average switchgrass yield falls within the range of similar traditional crops—in 2000 the average yields were 2.54 tons per acre for hay and 16.8 tons per acre for corn for silage.54 We assumed that there are no changes in crop distribution or total acreage between 2008 and 2030 and that the productivity of energy crops increases at the same rate as traditional crops.55 The energy crop harvest in 2030 is estimated to be 212.0 million dry tons in 2030 and 236.3 million dry tons in 2050. If the entire annual energy crop at the $50/ton price were converted to hydrogen, the resulting hydrogen energy would be 1.51.7 quads. We have assumed that the biomass actually available will be less than the potential maximum. Walsh, et al. estimated the size of the energy crop that would be available at a range of prices up to the maximum of $50/ton. 56 We have used the energy crop size at the $40/ton level, 35.2% of the maximum, for this study. We believe that the 35.2% availability reasonably represents the energy crop potential for hydrogen production when competing uses for the energy crop are considered, especially other power generation options such as co-firing with coal. Thus, the hydrogen available from energy crops in the years 2030 to 2050 ranges from 0.54 to 0.60 quads. The cost for biomass including 2.1.6.2 Agricultural Residues and Wood Waste
The quantities of agricultural residues and wood waste available for energy use at a range of price levels were estimated from the same modeling study by Walsh, et al. related to the dedicated energy crop analysis.57 Since some fraction of the residues must be left in the field to maintain the soil quality, the model assumed that only 30-40% of the residues could be removed from the field. The model considered only corn and wheat residues since other major crops such as soybeans provide very little residue and the harvest size of most other crops is small in 53
De La Torre Ugarte, D.G., et al. The Economic Impacts of Bioenergy Crop Production on U.S. Agriculture. July 2000, available at http://bioenergy.ornl.gov/pubs/econ_assess.html 54 USDA National Agricultural Statistics Service historical data, http://www.usda.gov/nass 55 The yield improvement for traditional crops was calculated from the USDA annual summary of yield for hay and corn for silage, which should be a similar type of farming to switchgrass. Since 1980 the average annual rate of yield improvement for those two crops was 0.55%. 56 Walsh, M.E., et al. Biomass Feedstock Availability in the United States: 1999 State Level Analysis. January 2000, available at http://bioenergy.ornl.gov/pubs/econ_assess.html 57 Ref. 56.
22
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
comparison to corn or wheat. The potential agricultural residue available at the highest price considered ($50/ton) is 151 million dry tons, with a hydrogen potential of 1.09 quads. Wood waste includes cuttings left in the forest after timber harvest, mill waste from lumber processing, and urban wood waste (construction and tree trimming). Currently, some wood waste is burned to provide steam and heat in the processing facility, so the processor must switch to a replacement energy source. The potential wood waste available at the highest price considered ($50/ton) is 135 million dry tons, with a hydrogen potential of 0.98 quads. Agricultural and wood wastes are not available at zero cost because there is significant cost incurred to collect, prepare, and deliver the materials to the facility where the conversion to hydrogen takes place. As a result, only 74% of the potential agricultural and wood wastes would be available at a cost of $36/ton, reducing the hydrogen available to 0.81 and 0.72 quads, respectively. 2.1.6.3 Municipal Solid Waste
In 1998, the average person in the U.S. generated 3.2 pounds per day of MSW, net of recycle.58 If this rate remains steady through the period 2030 to 2050, the total trash generated will increase linearly with the population. We have assumed that the hydrogen yield from MSW will be 58 kg/dry metric ton instead of 70 kg/dry metric ton because MSW includes a significant fraction of glass and metal that contains no recoverable hydrogen.59 The quantity of trash that would be generated in 2030 with population of 351.1 million is 205 million tons and in 2050 with population of 403.7 million is 236 million tons. 60 The hydrogen energy recoverable from these quantities of MSW is 1.22 to 1.40 quads. From the model of Walsh, et al., the projected MSW that would be recovered at a cost of $20/ton is 14.0% of the total in 2030 to 2050, or 29-33 million tons/year. The hydrogen yield from the reduced quantity of MSW is 0.17-0.20 quads. Our projections represent an upper limit for hydrogen from MSW, since a reduction in the per capita generation of trash is likely as landfill space is depleted and alternative uses for MSW such as electricity generation may be favored economically. 2.1.6.4 Landfill Gas
Municipal solid waste landfills generate gas due to anaerobic decomposition of organic matter in the trash. The landfill gas (LFG) contains on average about 50% methane that can be converted to hydrogen by steam reforming. The number and size of municipal landfills and the amount of LFG collected or flared per ton of waste in place (WIP) were
58
EPA Environmental Fact Sheet. Municipal Solid Waste Generation, Recycling and Disposal in the United States: Facts and Figures for 1998. April 2000, www.epa.gov/osw 59 Klass, D.L. Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San Diego, 1998, page 140. 60 A detailed explanation of the how the population projections are calculated is in Section 2.2.2
23
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
estimated from the EPA database for 1995.61 It is estimated that in 1995 landfills contained 5.8 billion tons of waste with potential gas generation of 158 scf/year/ton. If the current rate of waste disposal (3.2 lb/person/day) continues until 2030, the total amount of WIP will increase to 11.9 billion tons. For the purposes of our calculations, no new additions will be made to landfills starting in 2030 since all MSW will be converted to hydrogen directly (see Section 2.1.6.3). The total landfill gas generation is projected to be 953 billion scf/year. We have assumed that only 20% of the landfill gas, or 191 billion scf/year, will be recovered for hydrogen production, since currently some landfill gas is recovered for fuel or electricity generation and this practice is expected to become more widespread. The hydrogen potential from the steam reforming of 191 billion scf/year of methane is 0.13 quads. In practice, the calculated rate of landfill gas production can only be maintained for a limited number of years, since no additional MSW is accumulated, and in later years an increase in another renewable resource will be required to compensate. We have not accounted for the decline of landfill gas production since the contribution to the total quantity of hydrogen is relatively small, however, any focused analysis of landfill gas should address this issue. 2.1.6.5 Livestock Manure
The annual quantity of manure available nationwide was estimated by multiplying a “manure generation factor” for each type of livestock62 by the population of livestock in each state.63 We included only cattle, hogs, chickens, and turkeys since the populations of all other types of livestock are negligible in comparison. It is unlikely that all livestock manure could be recovered for hydrogen conversion. Like other renewable resources, this is more an issue of practicality and economics than technology. Only the manure generated in barns or other high-density locations would be available for capture and conversion to hydrogen. Klass64 includes a survey of potential for recovery of livestock manure as an energy source. The “collectible” fraction of manure ranges from 6.5% for beef cattle to 85% for hogs, with the other animals between the extremes. We have assumed that 20% of the manure generated annually will be collected for conversion to hydrogen, since it is reasonable to assume that much of the manure will continue to be used as fertilizer. The manure generation for each type of animal included is listed in Figure 2-1. A breakdown of manure generation by state is provided in Appendix B.
61
U.S. EPA database of municipal landfills, http://www.epa.gov/epaoswer/nonhw/muncpl/landfill/index.htm#list 62 Barker, J.C., et al. North Carolina Agricultural Chemicals Manual, 2002. 63 U.S. Dept. of Agriculture, livestock census 1997. 64 Klass (ref. 59), pp. 145-146.
24
Hydrogen from Renewable Energy Sources
Animal Type
Manure wet tons/yr/animal
Heifer Dairy cattle Beef cattle Hogs Chickens Turkeys
Directed Technologies Inc.
Population (thousands)
22.3 27,000 22.3 9,100 8.3 34,100 1.9 61,200 0.0355 1,464,500 0.112 98,950 Total Total @ 20% recovery for H2 production
Manure 10-6 × wet tons/year 602.1 202.9 283.0 116.3 52.0 11.1 1,267 253.5
Hydrogen Potential quads/year 0.70 0.24 0.33 0.13 0.06 0.01 1.47 0.29
Figure 2-1. Livestock manure generation by animal type. Wet manure is approximately 14.5% dry matter, and hydrogen yield is 7% by weight on a dry basis.
A separate estimate for the quantity of livestock manure generated annually in the U.S. from cattle, hogs, chickens, and turkeys is 1.36 billion tons (wet basis).65 Wet manure typically has a solids content of ~15%, so the dry matter available from the manure would be 204 million tons, with a theoretical hydrogen recovery of 1.26 quads. The close agreement between this number and the totals in Figure 2-1 give credibility to the stateby-state analysis. 2.1.6.6 Biomass Summary
As discussed in the introduction to Section 2.1.6 and quantified in the dedicated energy crop and agricultural/wood waste sections, competing uses for biomass feedstocks will reduce the quantity of biomass available for H2 generation. The estimated H2 from biomass of all types for transportation uses is 2.7 quads, or 63%, out of a potential total of 4.3 quads.66 The Vision for Bioenergy & Biobased Products in the United States prepared for the U.S. Congress sets goals for quantities of or relatively market share for biomass-derived products in the years through 2030.67 Although the goals set in the Vision are aggressive and exceed the biomass potential used in this analysis, the split among biopower, biobased transportation fuels, and biobased chemicals/materials is useful for comparison to the 63% for transportation uses estimated above. The Vision goal is the use of approximately 15.9 quads of biomass in 2030, of which 9.4 quads (59%) are projected for use as transportation fuels.68 The Vision fraction of biomass for transportation fuels is in close agreement to the 63% assumed in this report. A scenario in which the total biomass availability is used for H2 production was considered as an alternate case, and the detailed results for the alternate case are presented in Appendix A.
65
Doyle, Michael. Keeping Foodborne Pathogens Down on the Farm. Presentation available at http://www.fsis.usda.gov/OPPDE/animalprod/Presentations/KFPDF%20Aug%2001/ 66 The 4.3 quads of H2 corresponds to 9.6 quads of primary biomass energy. 67 Vision for Bioenergy & Biobased Products in the United States. October 2002. Available at http://www.bioproducts-bioenergy.gov/pdfs/BioVision_03_Web.pdf 68 This estimate of 59% of biomass resources for transportation requires the use of information sources in addition to the Vision report.
25
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.1.7 National Totals The quantity of hydrogen from renewable sources for 2030 and 2050 that we estimate to be technically achievable is listed in Figure 2-2. The state-by-state breakdown of hydrogen production is provided in Appendix B. H2 in 2030 (quads)
H2 in 2050 (quads)
Wind Solar Geothermal Biomass
22.90 5.9 0.43 2.66
22.90 5.9 0.43 2.76
Average H2 for 2040 (quads) 22.90 5.9 0.43 2.71 (4.3)a
Energy crops Agricultural residues Wood waste Municipal solid waste Landfill gas Livestock manure
0.54 0.81 0.72 0.17 0.13 0.29
0.60 0.81 0.72 0.20 0.13 0.29
0.57 0.81 0.72 0.18 0.13 0.29
Total Hydrogen
31.89
31.99
31.94 (33.5)a
a
The higher number reflects the potential if all available biomass were used for H2 rather than split with competing uses. Figure 2-2. Summary of hydrogen potential from renewable sources.
From the totals in Figure 2-2 it is clear that the production of 10 quads of hydrogen for transportation use from renewable energy sources is well within the technical limits of U.S. resources. The important factor in the development of a hydrogen-vehicle economy then becomes the cost of converting the renewable energy to hydrogen and distributing the hydrogen to the end user, topics that will be addressed in Sections 2.3, 2.4, and 3. 2.1.8 Other Sources of Renewable Energy for Hydrogen We have not included in this study any source of hydrogen that has the potential to generate less than 0.1 quads H2/year by 2050. The following renewable sources of H2 do not meet the minimum threshold and were excluded from the total availability listed in Figure 2-2. 2.1.8.1 Sewage Sludge
Conversion of sewage sludge, or biosolids, to energy may be an economically favorable method for using an otherwise wasted stream. However, the potential quantity of hydrogen from biosolids is small. Based on the biosolids generated by publicly owned treatment works (POTW’s) in 1998,69 the maximum hydrogen recoverable is 0.05 quads.
69
Biosolids Generation, Use, and Disposal in the United States. EPA Report EPA530-R-99-009, September 1999.
26
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.1.8.2 Biological Processes
There is a great deal of current research activity focused on producing hydrogen using bacteria or other simple organisms such as algae that are either naturally occurring or genetically modified. The research programs are too new to make even semi-quantitative predictions for hydrogen production potential in 2030. With current specific hydrogen production rates recently reported in the literature,70 the volume of bacterial culture necessary to generate enough hydrogen for a single fuel cell vehicle (~0.5 kg H2/day) is 77,000 liters. Several orders of magnitude improvements in specific yield are necessary before biological production of hydrogen would be developed at large scale. 2.1.9 Other Issues 2.1.9.1 Land Availability
Solar, wind, and biomass energy crops require substantial land areas to produce energy on the scale considered in this analysis. The land requirements are factored into the energy crop distribution and cost reported in Ref. 53 in Section 2.1.6. The Bureau of Land Management (BLM) is currently studying the potential for renewable energy production on federally-managed lands and has drafted a proposal for land use conditions and rental fees for wind energy development. The BLM proposal suggests royalty payments of 2% of revenues as a starting point for wind energy royalty payments. 2.1.9.2 Water Sources and Costs
Both electrolysis and biomass gasification/reforming require relatively large supplies of water. Electrolysis in particular requires ~2500 gallons of water per metric ton of hydrogen, as water impurities may result in contamination of the hydrogen and subsequent problems with fuel cell operation. We assume that the hydrogen generation processes will include a deionization step for final purification of potable water. Generating 10 quads of hydrogen from water electrolysis, assuming 100% conversion of the water, would require 207 billion gallons of water annually. To put the water needs in perspective, the U.S. potable water use is ~14,750 billion gallons per year, with industrial and agricultural non-potable water use adding ~133,000 billion gallons.71 For the U.S. as a whole, the marginal increase in water use for hydrogen generation would be negligible—only 0.16%. However, in some regions, particularly in California and the Southwest, the water supply is so limited that new sources of fresh water will be required to support the hydrogen infrastructure. The best possibility for expanding the water supply would be desalination of ocean water. Current desalination costs are approximately $2-$5 per 1000 gallons,72 which would add less than 2¢ per kilogram to the cost of hydrogen. This additional cost for water is unlikely to affect the economics of hydrogen production in any location near the coasts with the highest population density where the highest demand for hydrogen is likely to occur. 70
Freemantle, M. “Can We Exploit Hydrogenases?” Chemical & Engineering News, pp. 35-39, 22 July 2002. 71 U.S. EPA website, http://www.epa.gov/safewater/wot/howmuch.html 72 Ettouney, H.M., et al. “Evaluating the Economics of Desalination.” Chemical Engineering Progress, 98 12:32-39 (2002).
27
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.2 Geographic Distribution of Hydrogen Demand In this section we map the likely geographical distribution of hydrogen demand. Ideally, the supply would match demand in each state or county, leading to a trivial distribution plan. However, since supply and demand are generally not well matched geographically, the supply and demand are matched using a mathematical model described in Section 3. 2.2.1 Per Capita Gasoline Consumption We assumed that the demand for hydrogen in each state would be directly proportional to gasoline usage in that state since the 10 quads per year of hydrogen will displace gasoline as the primary passenger vehicle fuel. To estimate the gasoline (i.e., transportation energy) needs in 2030-2050, we assumed that the relative transportation energy need for each state scales with population, as determined by the ratio of population to gasoline consumption shown in Figure 2-3.73 We assumed that the relative gasoline use per capita remains the same for each state in the future. In general, the outliers from the linear relationship between population and gasoline consumption are states with high population concentration in urban areas with low vehicle ownership rates (e.g., New York) and states with relatively higher rural populations with high vehicle ownership rates (e.g., Texas). 2.2.2 State-by-State Population Projections We used the middle series projection from the Census Bureau for the years after the census in 2000.74 The Census Bureau projects state populations for the years 1995, 2005, 2015, and 2025 and national population for every year up to 2100. Growth rates (calculated as % population change over ten years) become fairly level by 2025. The growth rate in each state in the year 2025 was used to calculate the predicted state populations for year 2040. This estimate was then adjustedto the 2040 national population estimates by taking the predicted state population distribution (as % of national total) and multiplying by the 2040 national estimate. The time series (series A) state estimation and mid-level immigration national estimates were used.
73
The gasoline usage was taken from the 1999 DOE data http://eia.doe.gov/pub/state.data/pdf/summaries.pdf and the population from the 2000 Census. 74 U.S. Census Bureau population projections, http://www.census.gov/population/www/projections/natsum.html
28
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2,400 2,200
Gasoline Consumption (10^12 Btu/year)
2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
Population (thousands)
Figure 2-3. Correlation between gasoline usage and population by state (1999 data for gasoline usage, 2000 data for population).75
2.3 Capital Costs One of the benefits of renewable energy sources compared to conventional fossil fuels is that the “feedstock” is available at zero (wind and solar) or relatively low (biomass and geothermal) cost. As a result, the initial capital cost is the primary factor influencing the cost of energy from renewables. We have used current-day capital cost estimates for processes that are well-developed, such as biomass gasification, and future projections for emerging technologies such as wind and solar. The projections for wind, solar, and geothermal capital costs may be optimistic, as there is not a clear path to the projected costs for any of the three, with the possible exception of wind.
The capital costs used in this report are estimates or projections from publicly available sources or widely accepted correlations. We have not developed detailed designs for each of the renewable energy systems since the focus of this analysis was the overall hydrogen production and distribution pathway. We have attempted to avoid using the most optimistic or the most pessimistic scenario for projected costs. In Section 3.3 we examine the sensitivity of the cost of hydrogen to deviations in the capital costs presented here to quantity the effect of uncertain projections on the cost of hydrogen. For consistency, all of the capital costs were converted to end of year 2001 dollars by applying the Chemical Engineering Plant Cost Index (2001 = 394.3).76
75 76
See Ref. 73. Chemical Engineering, August 2002.
29
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.3.1 Wind Current wind turbine and balance of system costs are approximately $800-1,000/kWpeak. Estimates for future reductions in wind turbine costs vary, although the consensus is that reductions in cost will be gradual since wind is the most well-developed of the renewable electricity technologies. The DOE and EPRI in 1997 developed a roadmap for reducing costs for wind electricity generation.77 The 1997 projection for wind energy capital costs for a 50 MWpeak wind farm in 2030 was $635/kWpeak ($648/kWpeak in 2001 dollars). 2.3.2 Solar The current (2002) commercial solar prices are in the range of $5-$7 Wpeak at standard test conditions, depending on the type of cell and tracking system complexity. The NREL long-term goal for PV solar installations, including balance-of-plant items such as transformers, is $1/Wpeak. The progress of cost reduction for recent solar PV installations $16
$14
Installed System Cost ($/Wpeak)
$12
$10
$8
$6
$4
$2
$0 1980
1990
2000
2010
2020
2030
2040
2050
Year
Figure 2-4. The cost of photovoltaic solar systems (including solar module and balance of system components) by year of installation. The individual data points were taken from several reports from actual PV solar installations.78, 79, 80
is plotted in Figure 2-4. By fitting an exponential curve to the data in Figure 2-4, the capital cost for an installed PV solar system is projected to be $1-$1.50/Wpeak in the year 2040, so the NREL goal does not seem unreasonable as a projection for 2030 to 2050. The unknown factor that could dramatically reduce the solar cell cost (but not balance of system costs) is the development of relatively efficient and stable organic solar cells. The 77
EPRI and U.S. Department of Energy. Renewable Energy Technology Characterizations, TR-109496, December 1997. 78 Tomas, M.G., H.N. Post, and R. DeBlasio. Prog. Photovolt. Res. Appl., 7 (1999) 1-19. 79 Solar Electric Power Association. Large Systems Cost Report, 2001 Update, September 2001. 80 Personal communication with Dan Johns, BP Solar, 27 June 2002.
30
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
potential exists for organic solar cells to be manufactured using low-cost techniques such as screen or ink jet printing with inexpensive polymers. However, it is unlikely that organic solar cells will become efficient or stable enough by 2030 to significantly displace the conventional semiconductor cell technology for the widespread installations that will be needed to generate solar electricity on the quad-scale. 2.3.3 Geothermal The costs for geothermal-electric plants cover a wide range, from relatively low cost direct steam plants to relatively high cost binary systems. Most future geothermal facilities will be binary plants because of the relative abundance of low-temperature reservoirs. Meidav, et al. estimate that binary geothermal plants cost $2,500$4,000/kWpeak in 1994.81 We have projected (somewhat optimistically) that the capital cost for geothermal electric plants will decrease to $1,500/kWpeak in 2040 (2001 dollars). Since geothermal will contribute a maximum of 0.4 quads of H2, uncertainties in the capital cost estimate will not have a significant impact on the national distribution and cost of hydrogen. 2.3.4 Biomass to Hydrogen There are several biomass gasification processes that have been commercialized and tested on a pilot or an industrial scale. The technologies may be either oxygen/air gasification, steam gasification, or a combination of the two. The review by Klass provides a detailed breakdown of the subprocess costs for four well-known biomass gasification processes for methanol synthesis with 1650 dry metric tons/day biomass feed rate: Wright-Malta, Battelle, IGT, and Shell Oil.82 The hydrogen production capacity for a 1650 ton/day biomass gasification plant is ~115 metric tons/day, clearly a large-scale industrial gas facility rather than distributed production.
We included in the capital cost estimate only the subprocesses that would be required to produce hydrogen and chose the lowest-cost process, in this case the Battelle steam gasification process operating at atmospheric pressure. The costs for subprocess components for the Battelle process taken from Klass are listed in Figure 2-5. We are not endorsing or dismissing for technical reasons any of the aforementioned gasification processes. The only criterion for the selection is that we expect the cost of gasification processes in 2030-2050 to be at the low end of current designs, and the Battelle process is the lowest cost at the present time. There may be operational differences that favor one of the other processes in certain situations, for example, with a particular biomass feedstock.
81
Meidav, T. and J. Pigott. Geothermal Resources Council Bulletin, 23:10 (1994) 339-344. Klass, D.L. Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San Diego, 1998, p. 323. 82
31
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Process Component
Capital Cost, Thousand $
Feed Preparation
20,500
Gasifier
7,960
Reformer Feed Compressor
12,100
Reformer
17,100
Shift Reactors
10,300
Purification
16,000
Utilities/auxiliaries
33,400
Total
117,300
Figure 2-5. Biomass gasification capital costs for the atmospheric pressure Battelle steam gasification process, adapted from Reference 82. The costs have been adjusted to 2001 dollars.
2.3.5 Biomass-to-Electric The biomass-to-electric plant was assumed to be a 122 MWe low pressure indirectly heated gasifier with an advanced technology utility turbine and steam cycle, which was the lowest-capital cost option in a 1996 NREL study.83 We assumed that the overall plant efficiency will improve from ~35% currently to 45% in the future but that the equipment costs will not significantly change. The costs for the plant subsystems are listed in Figure 2-6, adjusted to 2001 dollars. Process Component
Capital Cost, Thousand $
Biomass feedstock handling and drying
10,859
Gasifier and gas cleanup
22,129
Gas turbine
19,682
Steam cycle
28,718
Balance of plant
56,587
Total
137,974
Figure 2-6. Biomass-to-electricity plant capital costs for a 122 MWe biomass gasification/combined cycle system. Adapted from Reference 83, adjusted to 2001 dollars.
We included only large-scale gasification electricity generation plants since the economies of scale allow for a reasonable cost of electricity. There may be some instances in which it is economically favorable for distributed electricity generation, for 83
Craig, K.R. and M.K. Mann. Cost and Performance Analysis of Biomass-Based Integrated Gasification Combined-Cycle (BIGCC) Power Systems. NREL/TP-430-21657, October 1996.
32
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
example, if the biomass transportation cost is more than we have assumed. However, we believe that the only way for biomass-derived electricity to be competitive with other renewable sources is at the large scale (see Section 3). 2.3.6 Landfill Gas Landfill gas is a special case since the “biomass” feedstock is methane gas. We have assumed that the methane is separated from the methane/carbon dioxide mixture using a conventional amine absorption process, then the purified methane is injected into the local natural gas distribution system. The quantity of methane generated by the landfill is assumed to be reformed to hydrogen at the vehicle fueling station. The methane purification sizing calculations used the median landfill size from the EPA Landfill Database.84 The capital cost for the amine absorption unit was estimated using cost factors for the main equipment (absorber and reboiler) and scaling factors for peripheral equipment, installation, and engineering costs. The methane reforming capital and operating costs were taken from previous estimates by Directed Technologies for distributed reforming of natural gas.85
For electricity generation using landfill gas as the fuel, the median size landfill produces enough gas to support a 5.56 MW turbine operating at 85% capacity factor. The capital cost for a gas turbine in the years after 2020 is projected to be $357 kW (corrected to 2001 dollars).86 The capital cost for the 5.56 MW turbine is projected to be approximately $2 million. 2.3.7 Hydrogen Pipelines There may be a need for either interstate/large capacity, local/small capacity hydrogen pipelines, or both, depending on the renewable energy source and distance from the user. There are hydrogen pipelines in existence now in geographic areas with extensive chemical processing and petroleum refining activities (e.g., Louisiana-Houston and Los Angeles). However, the current hydrogen pipeline network would be insufficient to support a nationwide fleet of fuel cell vehicles.
The natural gas transmission and distribution pipeline network provides a model on which to base the potential hydrogen distribution system. A rule-of-thumb estimate for the capital cost of large-scale hydrogen pipelines is a factor of 1.4-times more, on the basis of energy-transmitted, than equivalent natural gas pipelines.87 The construction of a natural gas pipeline between Wyoming and California, covering a distance of 717 miles, with a capacity of 840 billion Btu/day was recently approved.88 The cost of the pipeline project is expected to be $1.2 billion, a rate of $1.67 million per mile or $1.99/mile per 84
U.S. EPA database of municipal landfills, http://www.epa.gov/epaoswer/nonhw/muncpl/landfill/index.htm#list 85 Myers, D.B., et al. “Cost and Performance Comparison of Stationary Hydrogen Fueling Appliances”, Task 2 Report, April 2002, prepared by Directed Technologies Inc. for The Hydrogen Program Office, Office of Power Technologies, U.S. Department of Energy. 86 DiPietro, J.P. and J.S. Badin. “Technology Characterizations for the E3 (Energy, Economics, Environment) Pathway Analysis.” 1995 U.S. DOE Hydrogen Program Review, DE95009296. 87 See Ref. 12. 88 http://www.midamerican.com/html/press_release.asp?RELEASE_ID=136
33
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
million Btu/day of capacity. The new pipeline is actually an expansion beside an existing pipeline, so the cost for right-of-way is not as high as it could be for a completely new pipeline. The capital cost for a hydrogen pipeline was assumed to be 40% higher on a lower heating value basis than for a natural gas pipeline, for a cost factor of $2.79/mile per million Btu/day of capacity. The cost for local hydrogen distribution pipelines was calculated in a different way and is addressed in Section 2.4.4. 2.3.8 Interstate Electricity Transmission The capital cost for electricity transmission lines was calculated from a curve fit to data for existing above-ground transmission lines over the range 60-230 kV and 32-1,060 MW.89 The cost per mile of transmission line is reasonably well represented by a linear function of capacity, as shown in Figure 2-7, with only a minor dependence on voltage. The cost for a 1,000 MW transmission line is calculated to be a base cost of $102,000 per mile plus a marginal cost of approximately $500 per MW-mile. 900 800
Cost ($1000/mile)
700 600 500 400 300 200 100 0 0
200
400
600
800
1000
1200
Capacity (MW)
Figure 2-7. Cost of high-voltage electricity transmission lines as a function of capacity.
2.3.9 Electrolyzers The cost for electrolyzers was assumed to be the long-term DOE goal of $300/kW. All electrolysis was assumed to take place at local hydrogen dispensing stations, corresponding to a rated electrolyzer peak capacity of 1,700 kWth (920 kg/day) with 89
U.S. DOE, Energy Information Administration, www.eia.gov/cneat/pubs_html/feat_trans_capacity/table2.html
34
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
average hydrogen production of 1,173 kWth (630 kg/day). Large-scale electrolysis near the electricity source with hydrogen transport in pipelines (analogous to centralized biomass gasification) was considered as a pathway but was calculated to be economically unfavorable compared to distributed electrolysis (see Section 3.4.2). 2.3.10 H2 Filling Station (Compression, Storage, and Dispensing) All of the potential pathways to hydrogen from renewable sources require a method of dispensing the hydrogen to vehicles. The dispensing station that we included has a compressor rated to 10,000 psig; storage at 10,000 psig for 12 hours of average hydrogen sales; and dispensing equipment to vehicles with 5,000 psig tanks. The capital cost for 920 kg/day (peak) compression, storage, and dispensing module was taken from the Directed Technologies estimate for distributed natural gas reforming.90 2.4 Hydrogen and Electricity Cost Models The costs for renewable energy were estimated using discounted cash flow (DCF) analysis. The DCF analysis incorporates the cost of capital investment, raw materials, and operating and maintenance costs to calculate a levelized cost of hydrogen or electricity. The financial and operational assumptions are different for each step in the process of delivering hydrogen to automobiles, so the specific values for important variables are addressed in the subsections below. The detailed DCF calculations for all scenarios are included in Appendix C. The assumptions common to all the cost calculations are an inflation rate of 3% applied to fixed operating costs, marginal corporate income tax rate of 38% (federal and state combined), and corporate overhead of 15% of revenues.
No special tax preferences were assumed for renewable hydrogen generation in the years 2030-2050. Any special tax credits that exist now are expected to expire by the time renewables mature into a multi-quad per year market. 2.4.1 Renewable Electricity Generation The assumptions about plant lifetime and capital payback period are especially significant factors in the cost of hydrogen since the capital cost contributes the bulk of the cost of production. For example, wind is “free” once the wind turbine is constructed. We have assumed that in 2030-2050 electricity generation in small units and that the generation equipment is owned by independent or investor-owned power producers. The plant lifetime and payback period were both assumed to be 25 years. The DCF parameters used for renewable electricity generation are listed in Figure 2-8, and the DCF parameters used for electricity generation from biomass gasification are listed in Figure 2-9.
Electricity generation from landfill gas was handled differently since the generation method is assumed to be natural gas turbines at the landfill site with the electricity distributed through the grid. The DCF parameters used for electricity generation from biomass gasification are listed in Figure 2-10. 90
See Ref. 85.
35
Hydrogen from Renewable Energy Sources
Cost of capital Capital recovery period
Directed Technologies Inc.
Value
Reference
10.8%
CECA modeling91 Power producers are unregulated utilities or IPP Fraction of time the electricity generation is operational
25 years
Availability factor
98%
Capacity factor
varies
Land lease rate
2.5% of sales
Variable operating costs Fixed operating costs Operator labor Working capital
See Section 2.1 U.S. Bureau of Land Management draft memorandum
0.5¢/kWh 2% of initial investment per year 3 full-time (each 12 hrs/day @ $40/hour)
Estimate based on multi-megawatt plants, not counting maintenance or corporate/overhead staff
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-8. Discounted cash flow parameters for renewable electricity generation (solar, wind, geothermal). Value Cost of capital Capital recovery period
10.8%
Reference Ref. 91
25 years
Availability factor
95%
Capacity factor
85%
Fraction of time the electricity generation is operational Estimate
Biomass conversion
80%
Typical value; residual is inorganic ash
Plant thermal efficiency
45%
Feedstock costs
varies
Operation and maintenance Fixed operating costs Operator labor Working capital
See Section 2.1.6
3% of initial investment per year 1.5% of initial investment per year 12 full-time (each 12 hrs/day @ $40/hour)
Insurance 1%, property taxes 0.5% Estimate
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-9. Discounted cash flow parameters for biomass gasifier electricity generation.
91
U.S. Dept. of Energy. The Comprehensive Electricity Competition Act: A Comparison of Model Results. Report#:SR/OIAF/99-04. CECA is an acronym for the Comprehensive Electricity Competition Act.
36
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Value Cost of capital
10.8%
Capital recovery period
20 years
Capacity
5.56 MW
Reference Ref. 91
Availability factor
95%
Fraction of time the electricity generation is operational
Capacity factor
85%
Estimate
Biomass conversion
80%
Typical value; residual is inorganic ash
Plant thermal efficiency
50%
LHV basis
Feedstock costs Operation and maintenance Fixed operating costs Operator labor
Working capital
$1.644/Mscf natural gas
See Section 2.1.6.4
3% of initial investment per year 1.5% of initial investment per year 1 full-time (each 12 hrs/day @ $40/hour) 30 days sales accounts receivable 15 days expenses accounts payable
Insurance 1%, property taxes 0.5% Estimate
Figure 2-10. Discounted cash flow parameters for landfill gas electricity generation.
37
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.2 Electrolysis The electrolysis cost includes capital recovery and operating and maintenance, but the cost of electricity to operate the electrolyzer is accounted for separately (see Section 2.4.1 above). The DCF parameters used for electrolytic production of hydrogen are listed in Figure 2-11. No credit was taken for sales of the oxygen that is co-generated with the hydrogen, although it would be possible to improve the economics of the electrolysis process by selling the oxygen. Value Cost of capital Capital recovery period Capacity
10.8%
Ref. 91
10 years 920 kg/day
Availability factor
100%
Capacity factor
69%
Electrolyzer efficiency
76%
Water
Reference
$2/1000 gallons
Fraction of time the electrolyzer is operational Accounts for daily and seasonal fluctuations in demand LHV basis Typical value for desalination
Operation and maintenance
1.0% of initial investment per year
Fixed operating costs
1.5% of initial investment per year
Insurance 1%, property taxes 0.5%
Operator labor
1 full-time (each 12 hrs/day @ $20/hour)
Labor for dispensing station is assumed to be a lower-skill task than plant operation
Working capital
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-11. Discounted cash flow parameters for electrolyzer distributed hydrogen generation.
38
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.3 Biomass Gasification and Reforming to Hydrogen Biomass gasification and reforming to produce pure hydrogen is most economical at the industrial scale. The gasification plants are considered to be long-life plants with the same characteristics as an electricity plant. The DCF parameters used for biomass gasification and reforming to hydrogen are listed in Figure 2-12. Value Cost of capital Capital recovery period Capacity
10.8%
Reference Ref. 91
25 years 110 metric tons H2/day
Availability factor
95%
Capacity factor
85%
Fraction of time the electrolyzer is operational Accounts for daily and seasonal fluctuations in demand
Plant thermal efficiency
45.2%
LHV basis
Biomass feedstock
varies
See Section 2.1.6
Operation and maintenance
3% of initial investment per year
Fixed operating costs
1% of initial investment per year
Electricity usage Operator labor
Working capital
Insurance 0.5%, property taxes 0.5%
1 kW/(kg H2/hr) 12 full-time (each 12 hrs/day @ $40/hour)
Estimate
10 days biomass inventory 30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-12. Discounted cash flow parameters for biomass gasification and reforming to hydrogen in large-scale centralized plants.
39
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.4 Electricity Transmission Lines and Hydrogen Pipelines Electricity transmission lines and interstate hydrogen pipelines were assumed to be longterm assets that have a lower cost of capital (i.e., lower risk) and longer payback period than process equipment. The DCF parameters used for the interstate electricity transmission and hydrogen pipelines are listed in Figure 2-13. Value Cost of capital Capital recovery period Capacity Availability factor Capacity factor Operation and maintenance Fixed operating costs Compressor fuel (pipeline only)
Operator labor
Working capital
Reference
8% 25 years 1,000 MW (electricity); 332.3 metric tons H2/hour (pipeline) 100% Electricity 70%; Pipeline 60%
Fraction of time the system is operational Accounts for daily and seasonal fluctuations in demand
1% of initial investment per year 1.5% of initial investment per year
Insurance 1.0%, property taxes 0.5%
0.15% of throughput Electricity 1 full-time per 300 miles, Pipeline 2 full-time per 100 miles (each 12 hrs/day @ $40/hour)
Estimate
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-13. Discounted cash flow parameters for interstate electricity transmission and hydrogen pipelines.
The cost of local hydrogen distribution by pipeline is assumed to be the same as the difference between the city gate and commercial prices of natural gas. The markup for distribution beyond the city gate includes capital recovery and continuing operating costs for the natural gas distribution company. In 2000, the U.S. natural gas average price at the city gate was $4.62/Mscf and to commercial customers $6.59/Mscf.92 The difference of $1.97/Mscf ($0.0102/kWh) was used as the local pipeline distribution cost for hydrogen.
92
U.S. DOE, Energy Information Administration. Natural Gas Annual 2000.
40
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.5 Hydrogen Filling Stations (Compression, Storage, and Dispensing) All pathways for producing hydrogen from renewables include the final step of dispensing to vehicles. The cost for compression, storage, and dispensing modules were previously estimated by Directed Technologies Inc. in their analysis of distributed natural gas steam reforming.93 The DCF parameters used for the hydrogen filling station components are listed in Figure 2-14. Value Cost of capital Capital recovery period Capacity
Reference
10% 10 years 630 kg/day
Availability factor
100%
Fraction of time the station is operational
Capacity factor
69%
Accounts for daily and seasonal fluctuations in demand
Operation and maintenance
2% of initial investment per year
Fixed operating costs
1.5% of initial investment per year
Insurance 1.0%, property taxes 0.5%
Operator labor
1 full-time (each 12 hrs/day @ $40/hour)
Estimate
Corporate overhead
Working capital
0% of sales
Not included in filling station costs
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-14. Discounted cash flow parameters for hydrogen filling station compression, storage, and dispensing components.
2.4.6 Cost Summary The costs for each step in the hydrogen production chain are listed in Figure 2-15. The total cost of hydrogen for each pathway is calculated in Section 3.
93
See Ref. 85.
41
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Cost Electricity Generation (¢/kWh electric) Wind class 4 Wind class 5 Wind class 6 PV solar Geothermal Dedicated energy crops Ag residual & wood waste Livestock manure MSW Landfill gas
4.72 4.41 3.84 9.80 3.87 6.89 6.21 5.15 5.95 3.11
Electricity Transmission (¢/kWh electric/mile)
1.78 x 10-3
Electrolysis (¢/kWh H2 LHV)
3.90
Hydrogen Production (¢/kWh H2 LHV) Dedicated energy crops Ag & wood waste Livestock manure MSW Long-distance H2 pipeline (¢/kWh H2 LHV/mile)
5.25 5.05 3.95 4.34 7.2 x 10-4
Local H2 pipeline (¢/kWh H2 LHV)
1.02
Local compression, storage, and dispensing (¢/kWh H2 LHV)
1.89
Landfill gas-distributed steam reforming Landfill gas cleanup (¢/kWh natural gas LHV)
0.606
Local natural gas pipeline (¢/kWh natural gas LHV)
1.10
Long-distance natural gas pipeline (¢/kWh natural gas LHV/mile)
5.1 x 10-4
Distributed natural gas reforming (¢/kWh natural gas LHV)
4.10
Figure 2-15. Cost factors used to calculate the cost of hydrogen in Section 3.
42
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
3 Proposed Hydrogen Generation Network 3.1 Development of Hydrogen Distribution and Cost Model A simulation was created to generate a hydrogen production scenario for the continental U.S. based on resource availability and cost. The estimated costs of hydrogen are based on generation and transportation costs, with interstate distances taken into consideration. The model strives to simulate an unregulated hydrogen market in which consumers in each state buy hydrogen generated from the least expensive available resource from any other state. The algorithm takes as input the resource availability, base costs, and needs in each state and outputs the resulting distribution of hydrogen resource usage and prices.
In the model, hydrogen is purchased in small units (0.0001 quads), with each state purchasing the cheapest available hydrogen in each buying round until all of that state’s hydrogen needs are met or the resources are consumed. At the beginning of each round, a cost for hydrogen generated using each resource from each state to each state is calculated. If a particular resource in a state is non-existent or has been consumed, that resource is unavailable for purchase. The price of each resource is the sum of generation and transmission costs, with conversion efficiencies taken into consideration. For electricity-producing resources, electricity is transmitted from the generation location to the grid, and consumed at the individual refueling stations where hydrogen is generated via electrolysis. For biomass resources that are reformed, hydrogen is piped from the production center through a longdistance pipeline (when used out-of-state) and distributed via a local hydrogen pipeline to refueling stations. The cost of long-distance transmission of hydrogen and electricity is calculated based on center-to-center interstate distances. For in-state consumption of resources, only the cost of local transmission is included in the price. For biomass resources, generation costs include the costs of the feedstocks, their collection, and costs associated with gasifying and reforming to hydrogen. The cost of the hydrogen is calculated on a lower heating value (LHV) basis, and later converted to price per mass. Excluding the reformation of landfill gas, each resource is used to generate either electricity or hydrogen at a large-scale facility in the state in which the resource is located. For methods employing electrolysis, the following equation provides the cost of hydrogen (in $/kWh): 1 CH 2 = (C G + CT D) + C E + C CSD η e (1 − lT ) where η e is the electrolyzer efficiency (0.76) lT is transmission line loss (0.06) C G is the cost of electricity generation (varies) CT is the cost of electricity transmission per mile D is the interstate distance for transmission (varies)
43
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
C E is the electrolyzer cost C CSD is the cost of hydrogen compression, storage, and dispensing
For methods employing reformation of biomass (excluding landfill gas), the following equation is used to calculate the cost of hydrogen:
CH 2 =
1 C G + C P − L + C P − D D + C CSD (1 − l P ) D
where l P is pipeline loss of H2 per mile (zero) C G is the cost of hydrogen generation (varies) C P − L is the cost of transmitting hydrogen through a local pipeline C P − D is the cost per mile of transmitting hydrogen through a long-distance pipeline D is the interstate distance for transmission (varies) C CSD is the cost of hydrogen compression, storage, and dispensing We assume that landfill gas, unlike the other reformation methods, is purified to methane and pumped into the existing natural gas pipeline. Steam reformation units located at the refueling stations then remove an equivalent amount of natural gas from the pipeline to generate hydrogen. The cost of hydrogen from reformed landfill gas is then given by the following equation:
CH 2 =
1 (C C + C P − L + C P − D D) + C R + C CSD η R (1 − l P ) D
where η R is the reformer efficiency (0.70) l P is pipeline loss of natural gas per mile (zero) C C is the cost of landfill gas cleanup C P − L is the cost of transmitting natural gas through a local pipeline C P − D is the cost per mile of transmitting natural gas through a long-distance pipeline D is the interstate distance for transmission (varies) C R is the cost of reforming C CSD is the cost of hydrogen compression, storage, and dispensing The result of these cost models is that the cost of hydrogen depends on the method being used to generate it and the distance between the state with the resource and the state where the hydrogen is consumed. Figure 3-1 shows the cost of hydrogen produced from each method when consumed locally and the additional cost of transmitting the energy 500 miles (an arbitrary distance chosen to be representative for interstate transmission).
44
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
DE pyrolysis (A) ManureR
ManureE LFgasR
LFgasE
MSWE MSWR
WoodWasteR
AgriResR
3
DEcropsR
GeoT
4
AgriResE
DEcropsE
Solar WC6
5
WC5
6 WC4
cost of hydrogen, ($/kg)
7
WoodWasteE
zero mile transmission 500 mile transmission
8
DE pyrolysis (B)
Hydrogen costs do not include profit at any step94. Naturally, in the cases where multiple methods of hydrogen production exist for a single resource, the cheapest method will be selected. As a result, since gasification/reformation is the most cost effective method of hydrogen production from biomass, neither electrolysis nor pyrolysis from biomass resources were considered in the model. Figure 3-2 provides a breakdown of the cost for hydrogen from both the electrolysis and reformation methods used in the model.
2 1 0 method
Figure 3-1. Hydrogen costs from the various production methods. For biomass resources, E refers to electrolysis, R to reformation.
7
a
Electricity Electrolysis Comp.,Stor.,Disp. Transmission (per 500 miles)
3.5
3
H2 cost, $/kg
6
H2 cost, $/kg
Feedstock Gas. & Ref. Local H2 pipeline Local NG pipeline Comp.,Stor.,Disp. Pipeline (per 500 miles)
b
5 4
2.5
2
1.5
3
GeoT
Solar
WC6
WC5
WC4
LFgasR
0
ManureR
0
MSWR
0.5
WoodWasteR
1
AgriResR
1
DEcropsR
2
Figure 3-2. Cost breakdown for hydrogen from (a) renewable electricity and (b) reformed biomass.
94
With the possible exception of profits for transmission of electricity or hydrogen pipelines.
45
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
In each buying round during the simulation, each state has the opportunity to buy hydrogen from the method-state combination that provides the lowest cost. Because resources are limited, states will be forced to buy from a variety of methods and other states to meet their hydrogen needs. The result is a distribution of hydrogen costs that varies from state to state. States at longer distances from low-cost resources and states with larger hydrogen demand will pay more, on average, for hydrogen. If a resource in a particular state is depleted mid-round, the remaining states in that round are still able to purchase that resource. This is done in order to prevent a bias due to purchase order in the round and results in errors of no greater than a few hundredths of a quad for any single resource. As each state fulfills its H2 need, it stops buying resources. Rounds continue until all of the states have stopped buying. Due to the remoteness of Hawaii and Alaska, neither of these states were considered to be connected to the continental United States via pipeline or electrical grid. Together, Hawaii and Alaska represent roughly 0.6% percent of the national demand for hydrogen. As a result, these states are considered special cases requiring individualized, relatively small-scale hydrogen generation and transport. The model, therefore, focuses on the continental U.S., with Hawaii and Alaska neither consuming hydrogen nor providing resources for its production. 3.2 Results The simulation generates a round-by-round history of the hydrogen purchase choices for all states, providing potential resource preferences, locations, and transmission. The output of the simulation represents neither an optimized lowest-cost hydrogen scenario nor a plan for hydrogen production and distribution, but rather a representative hydrogen infrastructure that can provide insights into the variables affecting the cost of producing and distributing renewably-generated hydrogen. In order to evaluate the model results, we can look at the statistics for hydrogen production on a national basis as well as examining the results for individual states. 3.2.1 National Statistics Figure 3-3 and Figure 3-4 show the annual resource availability and consumption from the model. Consumption of each resource depends both on its availability and on the cost of hydrogen from that resource relative to all other resources. While biomass and geothermal resources represent a small hydrogen potential relative to solar and wind energy, these resources make significant contributions to the 10-quad national hydrogen demand due to their relatively low cost. Solar energy, however, makes no contribution since it is considerably more costly than the abundant and widespread Class 4 wind energy with which it must compete.
As stated earlier, the cost of hydrogen in a particular state depends, in part, on the distance between that state and the location of the resource. For two renewable resources with comparable costs, the addition of transmission costs (whether pipeline or electrical) may make a nearby but more expensive resource desirable over a cheaper but distant resource. This can be seen in Figure 3-3 where Class 5 and Class 6 wind are not completely consumed, yet large quantities of hydrogen are generated using the more
46
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
expensive Class 4 wind. This indicates that some high quality wind resources (Class 5 and 6) are stranded, being too remote from heavily populated states to be cost-efficient for hydrogen production. Biomass 2.7 Quads
a
Biomass 2.7 Quads
b
GeoT 0.43 Quads
Solar 5.9 Quads
WC4 5.3 Quads
WC4 18 Quads
GeoT 0.43 Quads
WC6 1.7 Quads
WC6 0.98 Quads
WC5 3.1 Quads WC5 0.48 Quads
Figure 3-3. Model resource availability and usage (in quads of hydrogen per year). (a) Resource availability, totaling 31.9 quads. (b) Resource usage, totaling 9.95 quads.
Resource
Abbrev.
Wind Class 4 Wind Class 5 Wind Class 6 Photovoltaic Solar Geothermal Dedicated Energy Agricultural Residue Wood Waste Municipal Solid Waste Livestock Manure Landfill Gas
WC4 WC5 WC6 Solar GeoT DEcrops AgriRes WoodWaste MSW Manure LFGas
Cost ($/kg) $4.54 $4.40 $4.13 $6.91 $4.14 $2.84 $2.77 $2.77 $2.53 $2.40 $2.70
Potential (quads/yr) 18.1 3.1 1.7 5.9 0.43 0.57 0.81 0.73 0.18 0.29 0.13
Usage (quads/yr) 5.3 0.48 0.98 0.0 0.43 0.57 0.81 0.74 0.18 0.29 0.13
Figure 3-4. Summary of hydrogen cost (w/ 500 mile transmission), availability, and usage for each resource predicted by the model.
The average cost of hydrogen in the U.S. is predicted to be $3.98/kg ($0.12/kWh). The large cost difference between hydrogen from renewable electricity and from biomass results in a multimodal hydrogen cost distribution, as shown in Figure 3-5, with a clear gap between these types of resources. The national minimum and maximum are $2.29/kg ($0.07/kWh) for manure and $5.12/kg ($0.15/kWh) for Class 4 Wind, respectively.
47
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2 1.8
H2 purchased (Quads/yr)
1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
2
2.5
3
3.5 4 Cost of hydrogen ($/kg)
4.5
5
5.5
Figure 3-5. National hydrogen cost distribution.
All else being equal, states with lower hydrogen demand are able to meet their needs at a lower average cost of hydrogen than states with large demands. This is because the model has each state purchasing the same quantity of hydrogen in each round. States with large hydrogen demands are forced to purchase quantities of expensive resources after small states have finished purchasing. This effect is displayed graphically in Figure 3-6, which demonstrates that the national average cost may decrease in consecutive rounds if states purchasing hydrogen at costs higher than the national average meet their needs and stop purchasing. Figure 3-7 provides the resulting average cost of hydrogen to each state. If, instead, all states were able to purchase the same fraction of their total demand in a purchase round, the average cost of hydrogen would be more uniform across states.
48
Directed Technologies Inc.
5
50
4.5
45
4
40
3.5
35
3
30
2.5
25
2
20
1.5
15
1
10
0.5
5
number of states purchasing
average cost of hydrogen ($/kg)
Hydrogen from Renewable Energy Sources
0 purchase round
Figure 3-6. Evolution of national average hydrogen cost throughout purchase process. Costs reflect hydrogen purchased in a given round and not the effect of hydrogen from previous rounds. AL AR AZ CA CO CT DE FL GA IA ID IL IN KS KY LA
$3.92 $3.72 $3.49 $4.09 $3.51 $3.25 $2.60 $4.72 $4.49 $3.11 $2.73 $4.16 $4.01 $3.23 $3.71 $3.63
MA MD ME MI MN MO MS MT NC ND NE NH NJ NM NV NY
$4.17 $4.11 $2.73 $4.17 $3.60 $3.84 $3.48 $2.60 $4.50 $2.54 $2.62 $2.72 $4.45 $3.08 $3.03 $4.52
OH OK OR PA RI SC SD TN TX UT VA VT WA WI WV WY
$4.30 $3.36 $3.33 $4.39 $2.62 $3.98 $2.56 $4.11 $4.03 $3.13 $4.34 $2.57 $3.68 $3.70 $2.64 $2.59
Figure 3-7. Average cost per kilogram of hydrogen for each state.
Generally, the least expensive hydrogen initially available to all states is from in-state biomass resources, followed by the biomass in neighboring states. Nearly half of all hydrogen from biomass is consumed in-state, with the amount transported out of state (via pipeline) decreasing with increasing interstate distance (see Figure 3-8). The average inter-state pipeline distance is 259 miles. As states consume their surrounding biomass resources, they are forced to purchase electricity for hydrogen generation, which may be transmitted over longer distances (see Figure 3-9). Only 24% of renewable electricity is used in-state for hydrogen production, and no obvious trend exists for the quantity of hydrogen transmitted as a function of transmission distance. The mean distance for renewable electricity transmission is 540 miles. It is interesting to note that
49
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
no resource is transmitted (via electrical lines or pipelines) more than 1500 miles, indicating that coastal states do not reach past the middle of the country to meet their hydrogen demand.
H2 piped (Quads/yr)
1.5
1.0
0.5
0
0
200
400
600
800 1000 distance (miles)
1200
1400
1600
1800
Figure 3-8. Distribution of transmission distances for hydrogen by pipeline. 1.8
H2 from electrolysis (Quads/yr)
1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0
0
500 1000 distance of electrical transmission (miles)
1500
Figure 3-9. Distribution of electricity transmission distances for electrolyzed hydrogen.
50
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
3.2.2 Selected State Statistics Because all forty-eight states in the model have access to the same resources, the combination of source states and resources any one state utilizes depends on the purchase decisions of all other states. Figure 3-10 through Figure 3-15 summarize the hydrogen purchases of four states: Texas, North Dakota, California, and New Jersey. These three states were selected as examples because they have very different demand and resource qualities. Figure 3-10 is a purchase history of the four states, showing the order of preference from available resources during the purchase process. Figure 3-11 summarizes the same data, emphasizing the source states and resources.
Texas has both a large hydrogen demand and a large supply of resources. It is able to meet a fraction of its demand with its own biomass resources and those of nearby states. However, as Figure 3-12a shows, the cost of hydrogen in Texas is dominated by its own wind resources. Texas supplies wind to other states as well, with the result that all of Texas’ Class 5 and 6 wind and most of Class 4 are consumed. North Dakota, like Texas, has significant wind resources. Unlike Texas, however, North Dakota has a low hydrogen demand and is distant from heavily populated states. North Dakota meets all of its hydrogen needs using biomass, both from within its borders and from other Midwestern and Plain states. As a result, the cost of hydrogen in North Dakota is comparatively low. According to the model, all of North Dakota’s wind is stranded, being too expensive for its own needs and too distant from states with high enough demands to justify the cost of transmission (see Figure 3-13b). For example, as Figure 3-10 and Figure 3-11 show, it is less costly for New Jersey to purchase Class 4 wind from as far as Minnesota than to purchase Class 5 wind from North Dakota (only one more state distant from New Jersey). California, like Texas, has both a large hydrogen supply and demand. In the purchase process, California continues to purchase hydrogen long after smaller states have met their hydrogen needs. This forces California to turn to resources which were too expensive for most other states, such as wind from Wyoming. As Figure 3-14 shows, the result is a lopsided distribution of hydrogen costs concentrated at more than $4.00/kg. The Eastern states have the disadvantage of having little local wind and, therefore, need to purchase wind electricity over long transmission distances for hydrogen production. New Jersey reaches as far as North Dakota for biomass resources before turning to the limited wind energy of Massachusetts, Maine, New York, New Hampshire, and Vermont (see Figure 3-10). This wind potential is quickly consumed, forcing New Jersey to purchase wind from Iowa and, finally, from Minnesota, which supplies more than half of New Jersey’s needs, making hydrogen relatively expensive (Figure 3-15a).
51
Hydrogen from Renewable Energy Sources
Source TX TX TX TX TX TX OK KS MO IA MN ND TX TX NM TX
TEXAS Resource quads/yr Manure 0.0096 MSW 0.0021 LFgas 0.0085 AgriRes 0.013 WoodWaste 0.0102 DEcrops 0.0126 DEcrops 0.0048 DEcrops 0.0016 DEcrops 0.0002 DEcrops 0.0007 DEcrops 0.0005 DEcrops 0.0002 WC6 0.0033 WC5 0.0331 WC6 0.0079 WC4 0.8655
$/kg $2.29 $2.42 $2.59 $2.65 $2.65 $2.72 $2.79 $2.83 $2.86 $2.91 $2.97 $2.98 $3.72 $3.99 $4.11 $4.13
Directed Technologies Inc.
Source CA CA CA CA CA OR WA NE NE NE SD KS ND CA CA NV CA CA UT NV NV WY
CALIFORNIA Resource quads/yr Manure 0.0059 MSW 0.0059 LFgas 0.0075 AgriRes 0.0036 WoodWaste 0.013 WoodWaste 0.0057 WoodWaste 0.0043 AgriRes 0.0024 WoodWaste 0.0001 DEcrops 0.0032 DEcrops 0.0088 DEcrops 0.002 DEcrops 0.0016 WC6 0.0245 GeoT 0.129 GeoT 0.192 WC5 0.0426 WC4 0.0619 GeoT 0.0419 WC5 0.0616 WC4 0.0633 WC6 0.5402
$/kg $2.29 $2.42 $2.59 $2.65 $2.65 $2.77 $2.83 $2.93 $2.93 $3.00 $3.01 $3.01 $3.02 $3.72 $3.73 $3.94 $3.99 $4.13 $4.15 $4.19 $4.34 $4.37
Source ND SD MN IA ND KS SD MN IA TX KS ND SD ND
NORTH DAKOTA Resource quads/yr Manure 0.0043 Manure 0.0013 Manure 0.0017 Manure 0.0001 MSW 0.0005 Manure 0.0002 MSW 0.0002 MSW 0.0009 MSW 0.0003 Manure 0.0001 MSW 0.0003 LFgas 0.0003 LFgas 0.0002 AgriRes 0.0152
$/kg $2.29 $2.33 $2.36 $2.41 $2.42 $2.44 $2.46 $2.49 $2.54 $2.55 $2.57 $2.59 $2.63 $2.65
Source NJ DE MD CT PA VA NC NJ MD PA VA NC OH MI IN TN NJ MD PA NJ NJ DE MD MD PA PA NJ VA NC MD SC VA KY WV IN IN GA OH IL IL KY IN TN IA IA IL MN WI MO IA MN ND MA ME MA NY NH VT PA MA NY NH ME ME IA MN
NEW JERSEY Resource quads/yr Manure 0.0003 Manure 0.0001 Manure 0.0004 Manure 0.0012 Manure 0.0006 Manure 0.0001 Manure 0.0007 MSW 0.0022 MSW 0.0004 MSW 0.0005 MSW 0.0003 MSW 0.0005 MSW 0.0003 MSW 0.0005 MSW 0.0002 MSW 0.0001 LFgas 0.0018 LFgas 0.0002 LFgas 0.0008 AgriRes 0.0002 WoodWaste 0.0009 AgriRes 0.0005 AgriRes 0.0013 WoodWaste 0.0011 AgriRes 0.0005 WoodWaste 0.003 DEcrops 0.0003 WoodWaste 0.0029 WoodWaste 0.006 DEcrops 0.0003 WoodWaste 0.0034 DEcrops 0.0001 WoodWaste 0.0011 DEcrops 0.001 AgriRes 0.0017 WoodWaste 0.0006 WoodWaste 0.0045 DEcrops 0.0003 AgriRes 0.0089 WoodWaste 0.0003 DEcrops 0.0003 DEcrops 0.0017 DEcrops 0.002 AgriRes 0.0045 WoodWaste 0.0002 DEcrops 0.0006 WoodWaste 0.0005 DEcrops 0.0021 DEcrops 0.0006 DEcrops 0.0007 DEcrops 0.0005 DEcrops 0.0002 WC6 0.0016 WC6 0.0007 WC5 0.0015 WC5 0.002 WC5 0.0005 WC5 0.0015 WC4 0.0061 WC4 0.0005 WC4 0.0036 WC4 0.0013 WC5 0.0007 WC4 0.0018 WC4 0.0414 WC4 0.146
$/kg $2.29 $2.31 $2.32 $2.32 $2.33 $2.35 $2.38 $2.42 $2.45 $2.46 $2.48 $2.51 $2.52 $2.56 $2.57 $2.58 $2.59 $2.62 $2.63 $2.65 $2.65 $2.67 $2.69 $2.69 $2.69 $2.69 $2.72 $2.72 $2.75 $2.75 $2.79 $2.79 $2.80 $2.80 $2.80 $2.80 $2.82 $2.82 $2.84 $2.84 $2.86 $2.87 $2.89 $2.90 $2.90 $2.91 $2.92 $2.93 $2.96 $2.97 $2.99 $3.06 $3.91 $4.12 $4.17 $4.23 $4.24 $4.24 $4.27 $4.32 $4.38 $4.38 $4.39 $4.53 $4.98 $5.06
Figure 3-10. Hydrogen purchases (in order of preference) for Texas, North Dakota, California, and New Jersey.
52
Hydrogen from Renewable Energy Sources
Source IA KS MN MO ND NM OK TX
TEXAS quads/yr Resource 0.0007 WC4 0.0016 WC5 0.0005 WC6 0.0002 DEcrops 0.0002 AgriRes 0.0079 WoodWaste 0.0048 Manure 0.9579 MSW LFgas
quads/yr 0.8655 0.0331 0.0112 0.0206 0.013 0.0102 0.0096 0.0021 0.0085
Directed Technologies Inc.
Source CA KS ND NE NV OR SD UT WA WY
CALIFORNIA quads/yr Resource 0.2939 WC4 0.002 WC5 0.0016 WC6 0.0057 GeoT 0.3169 DEcrops 0.0057 AgriRes 0.0088 WoodWaste 0.0419 Manure 0.0043 MSW 0.5402 LFgas
Source IA KS MN ND SD TX
NORTH DAKOTA quads/yr Resource 0.0004 AgriRes 0.0005 Manure 0.0026 MSW 0.0203 LFgas 0.0017 0.0001
quads/yr 0.1252 0.1042 0.5647 0.3629 0.0156 0.006 0.0231 0.0059 0.0059 0.0075
Source CT DE GA IA IL IN KY MA MD ME MI MN MO NC ND NH NJ NY OH PA SC TN VA VT WI WV
quads/yr 0.0152 0.0077 0.0022 0.0005
NEW JERSEY quads/yr Resource 0.0012 WC4 0.0006 WC5 0.0045 WC6 0.0468 DEcrops 0.0098 AgriRes 0.0042 WoodWaste 0.0014 Manure 0.0036 MSW 0.0037 LFgas 0.0032 0.0005 0.147 0.0006 0.0072 0.0002 0.0018 0.0057 0.0056 0.0006 0.0115 0.0034 0.0021 0.0034 0.0015 0.0021 0.001
quads/yr 0.2007 0.0062 0.0023 0.0107 0.0176 0.0245 0.0034 0.005 0.0028
Figure 3-11. Summary of source states and resources for hydrogen consumed by Texas, North Dakota, California, and New Jersey. 1.4
1.0 0.9
a
b
1.2
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
0.8 0.7 0.6
1
0.8
0.5
0.6
0.4 0.3
0.4
0.2
0.2 0.1
Manure
LFgas
MSW
4.2
WoodWaste
4
AgriRes
3.8
DEcrops
3.6
GeoT
3 3.2 3.4 Cost of hydrogen ($/kg)
Solar
2.8
WC6
2.6
WC5
2.4
WC4
0 2.2
Figure 3-12. Texas (a) hydrogen cost distribution, (b) resource usage. White bar shows unused hydrogen resource.
53
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
0.016
4.5
b
3.5
0.012
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
4
a
0.014
0.010
0.008
0.006
3
2.5
2
1.5
0.004
1
0.002
0
Manure
LFgas
MSW
WoodWaste
2.75
AgriRes
2.7
DEcrops
2.65
GeoT
2.45 2.5 2.55 2.6 Cost of hydrogen ($/kg)
Solar
2.4
WC6
2.35
WC5
2.3
WC4
0 2.25
0.5
Figure 3-13. North Dakota (a) hydrogen cost distribution, (b) resource usage. White bar shows unused hydrogen resource.
0.7
1
a
0.6
b
0.9
0.8
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
0.5
0.4
0.3
0.7
0.6
0.5
0.4
0.3
0.2
0.2 0.1
0.1 0
0
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
4.5
Solar
4
WC6
3 3.5 hydrogen cost ($/kg)
WC5
2.5
WC4
2
Figure 3-14. California (a) hydrogen cost distribution, (b) resource usage. White bar shows unused hydrogen resource.
54
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
0.20 5
0.18
a
x 10
-3
b
4.5
0.16 4
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
0.14
0.12
0.10
0.08
0.06
3.5
3
2.5
2
1.5
0.04 1
0.02 0.5
0
2
2.5
3
3.5
4
4.5
5
5.5
0
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Cost of hydrogen ($/kg)
Figure 3-15. New Jersey (a) hydrogen cost distribution, (b) resource usage. All resources were fully consumed.
3.3
Sensitivity Study for Important Variables
3.3.1 Hydrogen demand As we have seen, hydrogen produced by the gasification and reformation of biomass is significantly cheaper than hydrogen from electrolysis using renewable electricity. However, biomass resources are in short supply compared to the abundance of wind energy, particularly Class 4. The result is that hydrogen costs not only increase with increasing demand, but take a discontinuous jump once biomass resources are consumed. Figure 3-16 shows the national costs for hydrogen predicted by the model for national demands of 1, 3, 5, 10, and 15 quads/year while keeping costs and resource availability the same. As the demand increases, the cost of hydrogen becomes dominated by Class 4 wind. We can infer that hydrogen costs would take another jump when demand exceeded 26 quads/year, with Class 4 wind resources depleted and solar-produced hydrogen introduced to the market.
55
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
5.5
average H2 cost ($/kg)
5
maximum
4.5
4
mean
3.5
3
2.5
minimum 2
0
5
10 national H2 demand (Quads/yr)
15
Figure 3-16. Cost of hydrogen from renewables as a function of national demand for 1, 3, 5, 10, and 15 quads/year.
3.3.2 Low-cost solar Because Class 4 wind is so abundant and widespread, solar electricity must be costcompetitive with Class 4 wind electricity to play any significant role in the vehicular hydrogen market. The model was run for two low-cost solar scenarios: (1) the cost of solar electricity equals that of Class 4 wind, and (2) the cost of solar electricity equals that of Class 6 wind.
Scenario (1) represents a roughly 60% decrease in photovoltaic solar capital costs from the baseline model, or a 90% decrease from current costs. The result is that 0.71 quads/year of solar hydrogen are consumed, displacing some wind from all classes and some geothermal, and reducing the national average cost of hydrogen by two cents per kilogram to $3.96/kg. Figure 3-17a shows the national annual resource usage for this scenario. Scenario (2) represents a roughly 68% decrease in photovoltaic solar capital costs from the baseline model, or a 92% decrease from current costs. At this price, 2.1 quads/year of solar hydrogen are consumed, displacing wind and geothermal resources (see Figure 3-17b). The national average cost of hydrogen is reduced to $3.92/kg. Figure 3-18 summarizes the average price and solar resource usage from these alternate scenarios along with the baseline results.
56
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
a
Biomass 2.7 Quads
b
Biomass 2.7 Quads
WC4 4.1 Quads
WC4 5.2 Quads
GeoT 0.13 Quads
GeoT 0.39 Quads
Solar 0.71 Quads
WC6 0.44 Quads WC5 0.41 Quads
WC5 0.36 Quads WC6 0.44 Quads
Solar 2.1 Quads
Figure 3-17. Resource usage pies for scenarios in which solar electricity costs (a) equal wind Class 4 and (b) equal wind Class 6.
Cost of solar H2 w/ 500-mile transmission ($/kg)
$6.29 (baseline)
$3.92 (=WC4)
$3.51 (=WC6)
National average cost of H2 ($/kg)
$3.98
$3.96
$3.92
Solar H2 consumption (quads/year)
0.0
0.71
2.1
Figure 3-18. Results from scenarios with reduced costs for solar hydrogen.
3.3.3 Low-cost electrolyzed hydrogen The baseline model results reflect the high cost of hydrogen from electricity-generating resources compared to gasification/reformation of biomass. If renewable hydrogen from electrolysis could be made economically competitive with that from biomass, we might expect not only a significant decrease in the cost of hydrogen, but also a vastly different distribution for resource usage. Two scenarios were modeled in which hydrogen from Class 6 wind was competitive with biomass. Such a cost reduction would require both wind electricity and electrolysis costs to decrease, being unachievable through reductions in either component alone. The two scenarios examined assumed that, excluding transmission, hydrogen generated from Class 6 wind was at the same price as hydrogen from (1) dedicated energy crops and (2) livestock manure.95 The necessary cost reduction (in $/kg) to achieve these total costs for Class 6 wind were applied to the other electrolysis methods, as well.
95
The cost of hydrogen from Class 6 wind was actually made to be slightly more than expensive than the competing biomass, so that when choosing between biomass and Class 6 wind from in-state resources, the biomass was preferred.
57
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
In scenario (1), the national average cost of hydrogen was reduced to $3.25/kg from $3.98/kg. However, despite expectations, the national total consumption of each resource was relatively unchanged. The roughly 0.1 Quad/year increase in consumption of Class 6 Wind resources was taken from Class 4 Wind, indicating that previously stranded Class 6 wind resources had become economical compared to Class 4 resources. In scenario (2), the effect of scenario (1) was amplified, resulting in a national average cost for hydrogen of $2.91/kg. Figure 3-19 shows the resource utilization for both scenarios along with the baseline scenario. baseline
WC6 = DEcrops
WC6 = manure
20
a
Hydrogen from resource (Quads/yr)
18
b
c
16 14 12 10 8 6 4 2 0
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Figure 3-19. Resource uses and potential for three electrolyzed hydrogen costs. White bar shows unused hydrogen resource.
3.3.4 Low-cost electricity transmission As we have seen, the model predicts that significant Classes 5 and 6 wind resources are stranded due to the expense of electricity transmission. To investigate this effect, the model was run with two alternate electricity transmission costs: (1) 50% baseline transmission costs, and (2) 10% baseline costs.
Scenario (1) results in a national average hydrogen cost of $3.82/kg, while scenario (2) results in $3.64/kg. Figure 3-20 shows the national resource usage under each scenario, and Figure 3-21 shows summed Class 5 and 6 resource usage for states with those resources. baseline
50% transmission costs
10% transmission costs
20
a
Hydrogen from resource (Quads/yr)
18
b
c
16 14 12 10 8 6 4 2
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
WC4
0
Figure 3-20. National resource usage and potential for three electricity transmission costs. White bar shows unused hydrogen resource.
58
Hydrogen from Renewable Energy Sources
State MA ME NH VT NY VA WV NC TX NM CO UT MT SD ND WY CA AZ NV WA OR ID TOTAL
Potential Quads/yr 0.012 0.014 0.007 0.006 0.005 0.002 0.004 0.008 0.249 0.178 0.671 0.056 1.248 0.033 0.699 1.290 0.067 0.010 0.103 0.057 0.009 0.058 4.788
Directed Technologies Inc.
Resource Usage (Quads/yr) Baseline 50% reduced 90% reduced cost transmission transmission 0.013 0.013 0.013 0.015 0.015 0.015 0.007 0.008 0.008 0.007 0.007 0.007 0.006 0.006 0.006 0.002 0.002 0.002 0.004 0.004 0.005 0.009 0.008 0.008 0.250 0.250 0.253 0.114 0.178 0.179 0.000 0.393 0.672 0.042 0.056 0.057 0.042 0.271 1.248 0.033 0.034 0.034 0.000 0.699 0.700 0.656 1.290 1.291 0.067 0.067 0.067 0.010 0.010 0.010 0.103 0.103 0.103 0.057 0.057 0.057 0.009 0.009 0.009 0.017 0.059 0.059 1.463 3.538 4.803
Figure 3-21. Summed Class 5 and 6 wind resource usage for reduced electricity transmission costs. (Resource usages greater than potential are due to model rounding errors.)
The model predicts that reducing electricity transmission costs by 50% will “unstrand” all Class 6 wind and some Class 5. Reducing by 90% unstrands all Class 5 and 6 wind. Of the states with Class 5 and 6 wind resources, Montana has the second greatest potential, but is also the most stranded of all states. In no case does the reduced transmission cost cut into biomass usage. 3.4
Alternative Scenarios
3.4.1 Electrolysis at centralized facilities While not assessed in the model, it may be possible to perform electrolysis on a large scale at centralized facilities located near the resources. Hydrogen would then be transmitted via long-distance and local pipelines to the consumers at refueling stations. The advantages of such a scheme would be decreased capital and operating costs for the electrolyzer (due to increased capacity and capacity factor) as well as decreased electricity transmission distances and costs. However, the costs for local and longdistance pipelining would be the same as that for biomass gasification. Compression, storage, and dispensing costs would be roughly the same.
59
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
For a given reduction in electrolyser costs, further reductions in the cost of hydrogen are linear with the transmission distance since hydrogen pipelines are significantly cheaper (per unit of energy transmitted) than electric power lines. The mean distance of electricity transmission for renewable electricity in the model is 554 miles. At this distance and assuming negligible electricity transmission costs for centralized electrolysis, electrolyser costs (per unit of energy produced) would need to decrease by roughly 11% for centralized electrolysis to be competitive with the distributed electrolysis used in the model. Such reductions may be reasonable for large electrolysis facilities, making centralized electrolysis a likely player in an electrolyzed-hydrogen infrastructure. 3.4.2 Electrolysis at location of resource Another possible method for producing hydrogen from renewable electricity is to perform the electrolysis at the location of the resource (wind, solar, etc.) and to pipe the hydrogen gas to the refueling stations. The advantages of such a plan are that longdistance electricity transmission lines could be eliminated and that the national hydrogen infrastructure might be made independent of the electricity infrastructure.
In spite of the intuitive advantages, several costly disadvantages make the at-resource electrolysis strategy for most resources non-competitive compared to electricity transmission for electrolysis at the refueling station. First, the at-resource electrolyzer must be sized to meet the full capacity of the electricity generating system so that no renewable energy is wasted. Secondly, due to mismatched hydrogen production and demand rates, considerable hydrogen storage would be necessary near the production site.96 Lastly, while long-distance pipelines are cheaper than electricity transmission, local pipeline costs make piping the more expensive transmission method at most interstate distances. Figure 3-22, below, provides a cost comparison of at-resource versus at-station electrolysis using electricity generated by Class 6 wind. This analysis is generous in favor of at-resource electrolysis because the selection of Class 6 wind takes advantage of Class 6’s high capacity factor compared to other renewable electricity resources, (excluding geothermal). Lower capacity factors mean not only higher electricity costs, but also larger and more expensive electrolyzers for at-resource electrolysis. Furthermore, the analysis ignores the cost of additional hydrogen storage capacity required with at-resource electrolysis. Despite these advantages, at-resource electrolysis still comes out more expensive than at-station electrolysis, $4.77/kg vs. $4.13/kg. Like many hydrogen production methods, at-resource electrolysis may be desirable for niche applications where access to an electrical transmission infrastructure is limited. 96
For example, considering seasonal wind fluctuations in Amarillo, Texas, hydrogen storage capacity would need to accommodate at least 12% of the annual production. For perspective, 12% of annual hydrogen production for automotive use in 2040 (1.2 quads) would be 3 trillion scf of hydrogen, equal to the U.S. current underground storage capacity for natural gas. The minimum required storage capacity for an installation can be estimated by calculating the maximum value at any time of the integral of hydrogen production rate minus hydrogen consumption rate ( max( ∫ ( production − consumption) dt ) for all t ).
60
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
However, because Class 6 wind can be utilized more cost-effectively through at-station electrolysis, this method is more likely to make a significant contribution to meeting the national demand for hydrogen and is the only method considered further in this report. Of the resources examined in this report, only geothermal with its high capacity factor could possibly receive an economic advantage through at-source electrolysis. However, that advantage is estimated to be small even when extra storage costs are ignored, $3.85/kg vs. $4.14, or 7%. Cost component Electricity
Electrolyzer
Electricity Transmission Long-distance pipeline Local Pipeline Compression, Storage, Dispensing Total ($/kWh) Total ($/kg)
Electrolysis site At-resource At-station $0.051 $0.054
Comments
$0.060
$0.039
N/A
$0.012
Assuming 76% efficient electrolyzer. Atstation electricity penalized for 6% transmission line loss. At-resource electrolyzer cost for 49% capacity factor on electricity production and 0.6 engineering scaling factor. PAR = PAS*(0.49)-0.6 Cost provided for 500-mile transmission.
$0.004
N/A
Cost provided for 500-mile transmission.
$0.010 $0.019
N/A $0.019
$0.143 $4.77
$0.124 $4.13
While considerable extra storage would be necessary for at-resource H2 production, no penalty is applied here.
Figure 3-22. At-resource versus at-station electrolysis cost comparison (in $/kWh) for hydrogen production from Class 6 Wind.
3.4.3 Electricity from Renewables, Hydrogen from Natural Gas The goal of renewable electricity is presumably to decrease pollution and greenhouse gas emissions and to enhance energy security, rather than to provide the most economical source of electricity. All of the renewable resources investigated in this study can be used to generate electricity, although the costs of the production vary widely. Figure 3-23, below, shows that wind and geothermal resources produce electricity that is less costly than that from biomass. Here we will briefly investigate the relative economics and CO2 emissions from replacing some nuclear and fossil fueled electricity with wind and geothermal resources while producing hydrogen for transportation from natural gas.
61
Directed Technologies Inc.
Solar
Hydrogen from Renewable Energy Sources
ManureE
LFgasE
MSWE
WoodWasteE
GeoT
WC6
6
WC5
8 WC4
cost of electricity, (cents/kWh)
10
AgriResE
generated cost line loss 500 mile transmission DEcropsE
12
4
2
0 method Figure 3-23. Costs of renewably generated electricity.
Currently, 6-7% of the 12 quads of electricity consumed in the United States is generated from renewable sources. This percentage has held relatively steady for the past 50 years.97 The U.S. Department of Energy expects electricity consumption to rise by roughly 35% by 2025, but anticipates little change in the fraction supplied by renewables.98 If we take into consideration the 76% efficiency for electrolysis applied in this report, the 7 quads of hydrogen from geothermal and wind energy is equivalent to 9 quads of electricity. This quantity may be as much as one third of the annual electricity consumption by the year 2040. It is doubtful whether the United States electrical infrastructure could support this high fraction of renewable energy without grid stability problems. However, for the sake of this brief analysis, we will assume that 9 quads/yr of renewable electricity is possible. The estimated national annual cost and CO2 emissions for this scenario is compared to the model baseline in Figure 3-24. Based on the prices of electricity used in this report (projected for 2040) and an average wholesale cost for electricity of $0.03/kWh from non-renewable production methods,99 9 quads/yr of renewable electricity is estimated to 97
Annual Energy Review 2001: Energy Overview: 1949-2001. Energy Information Administration. http://www.eia.doe.gov/emeu/aer/pdf/pages/sec8_5.pdf 98 Annual Energy Outlook 2003 with projections to 2025. Energy Information Administration. http://www.eia.doe.gov/oiaf/aeo/pdf/appa.pdf 99 This cost is generously low to account for advances in conventional electricity generation.
62
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
cost the United States an additional $40billion annually and to alleviate roughly 1500Mtonnes/yr100 of CO2 emissions generated from the current mix of coal, natural gas, and nuclear electricity. The average cost of hydrogen from renewable electricity resources in the year 2040 was estimated to be $4.49/kg, or $280 billion per year for 7 quads. The estimated cost for the same hydrogen reformed from natural gas is estimated to be roughly $120 billion annually, with a CO2 release of 600 Mtonnes/year101. The result of this alternate scenario is roughly 900Mtonnes/year additional CO2 reduction for $235 billion/year less when compared to the model baseline. (For perspective, total U.S. CO2 emissions from all human sources in 2000 were estimated at 5800 Mtonnes.102) Scenario
BASELINE: Renewable H2; Fossil-based elec. 7 quads/yr renewable elec.; NG-based H2
Cost of 9 quads electricity compared to baseline ($/yr)
$0 (from coal, natural gas, nuclear) $40billion (from wind and geothermal)
Cost of 7 quads hydrogen ($/yr)
$395 billion (from wind & geothermal) $120 billion (from steam methane reformation)
CO2 emissions compared to baseline (Mtonnes/yr)
0 -1500 + 600 = -900
Figure 3-24. Estimated cost and CO2 emissions of alternative renewable electricity and hydrogen from natural gas compared to the model baseline.
In reality, each renewable energy source has its own merits and disadvantages for each application, considering storage, stability, and distribution. However, we generalize that if faced with a choice between using a quantity of renewable electricity as grid electricity or hydrogen, it is both less expensive and less greenhouse gas intensive to generate electricity with renewables and produce hydrogen from fossil fuels than vice-versa. 3.4.4 Nuclear Hydrogen While not addressed in the model, methods for producing hydrogen at nuclear plants have been proposed. Nuclear power has the advantages of zero greenhouse gas emissions, relatively low cost compared to renewable electricity (at least in the near-term), and excess heat. Because the electrical efficiency of electrolysis improves at elevated temperatures, nuclear’s excess heat provides a lower cost, higher efficiency method for electrolytic production of hydrogen.
Another option (besides electrolysis) for nuclear production of hydrogen is thermochemical processing, in which water is split into hydrogen and oxygen via a set of 100
Based on CO2 emissions of 245, 134, and zero Mtonnes/quad of electricity and a 50:35:15 mix for coal, natural gas, and nuclear electricity, respectively. 101 Based on a 70% efficient steam methane reforming process producing hydrogen for $2/kg. 102 Emissions of Greenhouse Gases in the United States 2000. Energy Information Administration. November 2001.
63
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
chemical reactions with heat supplying all of the energy. One such process developed by General Atomics produces H2SO4 and HI as intermediary species in the process, with sulfur and iodine recycled in the process. Thermal efficiencies of roughly 50% are achievable103, but little or no electricity is consumed in the process. 3.5 Study Limitations The results reported above are based on reasonable estimates of capital and operating costs taken from the literature. We did not seek to optimize each source of hydrogen nor did we seek to design new processes. Some of the technology and cost projections are speculative since predictions about costs forty years from now have a high degree of uncertainty.
We started with the assumption that 10 quads per year of hydrogen would be used for transportation purposes and that all of that hydrogen would be derived from renewable resources. Those assumptions likely resulted in a sub-optimal solution for the U.S. energy market as a whole, especially with respect to the cost of hydrogen. For example, natural gas may be used to produce low-cost hydrogen, freeing up renewable resources for residential, commercial, and industrial electricity (see Section 3.4.3). The study assumed that the 10 quads of vehicular hydrogen from renewable resources would not need to compete with hydrogen from non-renewable sources. Non-renewable hydrogen would likely be produced from either the steam methane reformation (SMR) of natural gas or from electrolysis using electricity produced primarily from coal, natural gas, and nuclear energy. Figure 3-25 compares hydrogen costs for renewable hydrogen with similar processes using non-renewable resources. Reformation of natural gas for hydrogen production may take place at large plants or at refueling station installations, and representative costs for each are included in the figure. We assume that electrolysis always takes place at the refueling station. A given process using a non-renewable energy source is generally less expensive than its renewable alternative, but reformation is considerably less expensive than electrolysis (~25-50%), regardless of the energy sources used. Had renewable hydrogen been required to compete with some quantity and distribution of non-renewable sources, a different infrastructure with a lower average cost of hydrogen could be expected.
103
Schultz, Ken. Efficient Production of Hydrogen from Nuclear Energy (presentation). California Hydrogen Business Council. June 27, 2002.
64
Hydrogen from Renewable Energy Sources
H2 Production Method
Reformation processes
Electrolysis processes
Directed Technologies Inc.
Cost Range ($/kg H2)
Comments
Large plant SMR
$2.21 $2.63
Based on existing 327 and 33 tonne/day facilities running on natural gas w/ 500mile hydrogen pipeline and natural gas cost of $4.17/106 Btu (average of industrial and utility prices in 2000)
Station-size SMR
$1.99 $3.38
Based on proposed 920 and 115 kg/day installations running on natural gas
Large plant biomass
$2.40 $2.84
Conventional electricity
$3.74 $4.67
Based on proposed 108 tonne/day facilities running on biomass resources w/ 500-mile hydrogen pipeline Based on electricity at wholesale costs of 3-6 cents/kWh w/ 500-mile transmission
Renewable electricity
$4.13 $4.54
Based on projected costs for geothermal and wind electricity w/ 500-mile transmission
Figure 3-25. Hydrogen costs from renewable and non-renewable sources, including transmission, compression, storage, and dispensing costs.
65
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
4 Conclusions The annual generation of 10 quads of hydrogen in the years 2030-2050 from renewable sources for transportation uses in the U.S. is technically achievable and, according to our model, leads to a national average hydrogen cost of $3.98/kg ($33.24/GJ, LHV basis), excluding profits. If hydrogen fuel cell vehicles achieve the projected efficiency (2.2times higher than conventional internal combustion vehicles), the cost of hydrogen would be equivalent to $1.81 per gallon for gasoline untaxed and without profit for the refiner or the distributor.104 The state average costs of renewable hydrogen vary widely ($2.54$4.72/kg), depending on the availability of resources and the necessary transmission distances. Wind and biomass are the most significant resources (on an energy supplied basis) for hydrogen production, with geothermal playing a small role due to its limited potential. A significant quantity of high-quality wind resources (Classes 5 and 6) in the central U.S. will be stranded due to the prohibitive cost of transmission. Hydrogen from renewable electricity is expensive compared to that from the reformation of biomass for three important reasons. First, most renewable electricity, especially wind and solar, is expensive compared to electricity from fossil fuels due primarily to high capital costs for the wind and solar installations with relatively low capacity factors (all
Prepared by:
Duane B. Myers, Gregory D. Ariff, Brian D. James, & Reed C. Kuhn
One Virginia Square 3601 Wilson Boulevard, Suite 650 Arlington, Virginia 22201 703/243-3383
Prepared for:
The Hydrogen Program Office Office of Power Technologies U.S Department of Energy Washington, D.C. Under Grant No. DE-FG01-99EE35099
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
FORWARD This work was funded by the Hydrogen Program Office of the U.S. Department of Energy under Grant No. DE-FG01-99EE35099 and represents the third and final task of three to be completed under this contract. The first task presented a broad overview of the costs for creating infrastructures to supply direct hydrogen, methanol, and gasoline to support fuel cell vehicles (FCV’s). A conclusion of the report resulting from the first task was that “the costs of maintaining the existing gasoline infrastructure per vehicle supported are up to two times more expensive than the estimated costs of maintaining either a methanol or a hydrogen fuel infrastructure”. The second task was to provide a detailed analysis of the cost of providing small-scale stationary hydrogen fueling appliances (HFA’s) for the on-site production and storage of hydrogen from natural gas to fuel hydrogen FCV’s. Four potential reforming systems were studied: 10-atmosphere steam methane reforming (SMR) with pressure-swing adsorption (PSA) as gas cleanup, 20-atm SMR with metal membrane gas cleanup, 10-atm autothermal reforming (ATR) with PSA gas cleanup, and 20-atm ATR with metal membrane gas cleanup. The third task, the subject of this report, was to identify a pathway for producing 10 quads of hydrogen for transportation uses per year from renewable sources in the years 2030 to 2050 (1 quad = 1015 Btu). The 10 quads of hydrogen is approximately the quantity that would be needed if all passenger vehicles were converted to hydrogen fuel cell vehicles. The scope of the analysis was limited to those categories of renewable sources that are potentially significant contributors on a 10 quad scale (wind, solar, geothermal, and biomass). We acknowledge the support of the Department of Energy, especially Sig Gronich, DOE Hydrogen Team Leader.
i
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
TABLE OF CONTENTS FORWARD ........................................................................................................................ i LIST OF FIGURES ......................................................................................................... iv EXECUTIVE SUMMARY .............................................................................................. 1 1 Basic Information on Renewable Energy Sources................................................. 6 1.1 What Does 10 Quads of Hydrogen Represent?.................................................. 6 1.1.1 Vehicle Energy Use ........................................................................................ 6 1.1.2 Natural Gas Pipeline Capacity ........................................................................ 6 1.1.3 Electricity Consumption and Grid Capacity ................................................... 7 1.2 Renewable Energy Sources for Hydrogen.......................................................... 7 1.2.1 Wind................................................................................................................ 8 1.2.2 Solar .............................................................................................................. 10 1.2.3 Geothermal.................................................................................................... 14 1.2.4 Biomass......................................................................................................... 15 2 Hydrogen Supply, Demand, and Economics ........................................................ 17 2.1 Renewable Energy Resources (Supply) ............................................................ 17 2.1.1 Competition for Renewable Energy Resources ............................................ 17 2.1.2 Electricity Grid Considerations..................................................................... 17 2.1.3 Wind.............................................................................................................. 18 2.1.4 Solar .............................................................................................................. 20 2.1.5 Geothermal.................................................................................................... 20 2.1.6 Biomass......................................................................................................... 21 2.1.7 National Totals.............................................................................................. 26 2.1.8 Other Sources of Renewable Energy for Hydrogen ..................................... 26 2.1.9 Other Issues................................................................................................... 27 2.2 Geographic Distribution of Hydrogen Demand............................................... 28 2.2.1 Per Capita Gasoline Consumption ................................................................ 28 2.2.2 State-by-State Population Projections........................................................... 28 2.3 Capital Costs........................................................................................................ 29 2.3.1 Wind.............................................................................................................. 30 2.3.2 Solar .............................................................................................................. 30 2.3.3 Geothermal.................................................................................................... 31 2.3.4 Biomass to Hydrogen.................................................................................... 31 2.3.5 Biomass-to-Electric....................................................................................... 32 2.3.6 Landfill Gas .................................................................................................. 33 2.3.7 Hydrogen Pipelines....................................................................................... 33 2.3.8 Interstate Electricity Transmission ............................................................... 34 2.3.9 Electrolyzers ................................................................................................. 34 2.3.10 H2 Filling Station (Compression, Storage, and Dispensing)......................... 35 2.4 Hydrogen and Electricity Cost Models ............................................................. 35 2.4.1 Renewable Electricity Generation ................................................................ 35
ii
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.2 Electrolysis.................................................................................................... 38 2.4.3 Biomass Gasification and Reforming to Hydrogen ...................................... 39 2.4.4 Electricity Transmission Lines and Hydrogen Pipelines .............................. 40 2.4.5 Hydrogen Filling Stations (Compression, Storage, and Dispensing) ........... 41 2.4.6 Cost Summary............................................................................................... 41 3 Proposed Hydrogen Generation Network ............................................................ 43 3.1
Development of Hydrogen Distribution and Cost Model................................ 43
3.2 Results .................................................................................................................. 46 3.2.1 National Statistics ......................................................................................... 46 3.2.2 Selected State Statistics................................................................................. 51 3.3 Sensitivity Study for Important Variables ....................................................... 55 3.3.1 Hydrogen demand......................................................................................... 55 3.3.2 Low-cost solar............................................................................................... 56 3.3.3 Low-cost electrolyzed hydrogen................................................................... 57 3.3.4 Low-cost electricity transmission ................................................................. 58 3.4 Alternative Scenarios.......................................................................................... 59 3.4.1 Electrolysis at centralized facilities .............................................................. 59 3.4.2 Electrolysis at location of resource ............................................................... 60 3.4.3 Electricity from Renewables, Hydrogen from Natural Gas.......................... 61 3.4.4 Nuclear Hydrogen......................................................................................... 63 3.5 Study Limitations................................................................................................ 64 4 Conclusions.............................................................................................................. 66 APPENDIX A: HIGHER BIOMASS AVAILABILITY ESTIMATES..................... 67 APPENDIX B: MODEL INPUT - RENEWABLE ENERGY RESOURCES BY STATE IN 2040 ............................................................................................................... 70 APPENDIX C: MODEL OUTPUT - STATE BUYING HISTORIES FOR BASELINE ...................................................................................................................... 71
iii
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
LIST OF FIGURES Figure 1-1. Classes of Wind Power Density...................................................................... 9 Figure 1-2. Wind Class Distribution in the United States. (Source: Wind Energy Resource Atlas of the United States, available at http://rredc.nrel.gov/wind/pubs/atlas/maps/chap2/2-01m.html)............................... 10 Figure 1-3. Average solar radiation during the month of June in the U.S. (units are kWh/day-m2)............................................................................................................. 11 Figure 1-4. Solar radiation striking a fixed flat plate tilted toward the south at the latitude angle for selected U.S. cities..................................................................................... 11 Figure 1-5. PV solar-electric examples............................................................................ 13 Figure 2-1. Livestock manure generation by animal type. Wet manure is approximately 14.5% dry matter, and hydrogen yield is 7% by weight on a dry basis.................... 25 Figure 2-2. Summary of hydrogen potential from renewable sources............................. 26 Figure 2-3. Correlation between gasoline usage and population by state (1999 data for gasoline usage, 2000 data for population). ............................................................... 29 Figure 2-4. The cost of photovoltaic solar systems (including solar module and balance of system components) by year of installation. The individual data points were taken from several reports from actual PV solar installations., , .............................. 30 Figure 2-5. Biomass gasification capital costs for the atmospheric pressure Battelle steam gasification process, adapted from Reference 82. The costs have been adjusted to 2001 dollars. ........................................................................................... 32 Figure 2-6. Biomass-to-electricity plant capital costs for a 122 MWe biomass gasification/combined cycle system. Adapted from Reference 83, adjusted to 2001 dollars........................................................................................................................ 32 Figure 2-7. Cost of high-voltage electricity transmission lines as a function of capacity. ................................................................................................................................... 34 Figure 2-8. Discounted cash flow parameters for renewable electricity generation (solar, wind, geothermal). .................................................................................................... 36 Figure 2-9. Discounted cash flow parameters for biomass gasifier electricity generation. ................................................................................................................................... 36 Figure 2-10. Discounted cash flow parameters for landfill gas electricity generation. .... 37 Figure 2-11. Discounted cash flow parameters for electrolyzer distributed hydrogen generation.................................................................................................................. 38 Figure 2-12. Discounted cash flow parameters for biomass gasification and reforming to hydrogen in large-scale centralized plants................................................................ 39 Figure 2-13. Discounted cash flow parameters for interstate electricity transmission and hydrogen pipelines. ................................................................................................... 40 Figure 2-14. Discounted cash flow parameters for hydrogen filling station compression, storage, and dispensing components......................................................................... 41 Figure 2-15. Cost factors used to calculate the cost of hydrogen in Section 3. ............... 42 Figure 3-1. Hydrogen costs from the various production methods. ................................ 45 Figure 3-2. Cost breakdown for hydrogen from (a) renewable electricity and (b) reformed biomass...................................................................................................... 45 Figure 3-3. Model resource availability and usage (in quads of hydrogen per year). ..... 47 Figure 3-4. Summary of hydrogen cost (w/ 500 mile transmission), availability, and usage for each resource predicted by the model. ...................................................... 47
iv
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Figure 3-5. National hydrogen cost distribution. ............................................................. 48 Figure 3-6. Evolution of national average hydrogen cost throughout purchase process. Costs reflect hydrogen purchased in a given round and not the effect of hydrogen from previous rounds. ............................................................................................... 49 Figure 3-7. Average cost per kilogram of hydrogen for each state. ................................ 49 Figure 3-8. Distribution of transmission distances for hydrogen by pipeline.................. 50 Figure 3-9. Distribution of electricity transmission distances for electrolyzed hydrogen. ................................................................................................................................... 50 Figure 3-10. Hydrogen purchases (in order of preference) for Texas, North Dakota, California, and New Jersey. ...................................................................................... 52 Figure 3-11. Summary of source states and resources for hydrogen consumed by Texas, North Dakota, California, and New Jersey. .............................................................. 53 Figure 3-12. Texas (a) hydrogen cost distribution, (b) resource usage. .......................... 53 Figure 3-13. North Dakota (a) hydrogen cost distribution, (b) resource usage. .............. 54 Figure 3-14. California (a) hydrogen cost distribution, (b) resource usage..................... 54 Figure 3-15. New Jersey (a) hydrogen cost distribution, (b) resource usage. ................. 55 Figure 3-16. Cost of hydrogen from renewables as a function of national demand for 1, 3, 5, 10, and 15 quads/year. ...................................................................................... 56 Figure 3-17. Resource usage pies for scenarios in which solar electricity costs (a) equal wind Class 4 and (b) equal wind Class 6. ................................................................. 57 Figure 3-18. Results from scenarios with reduced costs for solar hydrogen. .................. 57 Figure 3-19. Resource uses and potential for three electrolyzed hydrogen costs........... 58 Figure 3-20. National resource usage and potential for three electricity transmission costs........................................................................................................................... 58 Figure 3-21. Summed Class 5 and 6 wind resource usage for reduced electricity transmission costs. .................................................................................................... 59 Figure 3-22. At-resource versus at-station electrolysis cost comparison (in $/kWh) for hydrogen production from Class 6 Wind.................................................................. 61 Figure 3-23. Costs of renewably generated electricity. ................................................... 62 Figure 3-24. Estimated cost and CO2 emissions of alternative renewable electricity and hydrogen from natural gas compared to the model baseline. ................................... 63 Figure 3-25. Hydrogen costs from renewable and non-renewable sources, including transmission, compression, storage, and dispensing costs........................................ 65
v
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
EXECUTIVE SUMMARY Introduction This report is the final in a series of studies by Directed Technologies, Inc. of the cost and infrastructure requirements to supply hydrogen for fuel cell vehicles (FCVs). The previous studies have concentrated on the early development of the hydrogen infrastructure to supply a limited number of FCVs. In this study, the focus shifts to future years in which a significant fraction of passenger cars and trucks are FCVs. The goal of the current analysis was to develop a technically feasible pathway to supply 10 quads1 per year of hydrogen fuel from renewable energy sources for transportation uses in the years 2030 to 2050. The midpoint year 2040 was used for calculations of resource availability and demand that may change over the 20 year period. To put 10 quads of hydrogen energy in perspective, if the passenger vehicles currently on the road were converted to FCVs, 10 quads of hydrogen would be sufficient to fuel all of those vehicles.2 Approach and Results The first task was to determine the total potential hydrogen generation from renewable resources in the U.S. in 2030-2050 to ensure that the production of 10 quads of hydrogen was possible. All renewable resources were considered; however, only the resources that could make a contribution of at least 0.1 quads per year were included: biomass, solar, wind, and geothermal. Hydrogen from the biomass resources was produced through gasification and steam reforming at centralized facilities with pipeline delivery to the service stations. Hydrogen from the electricity-generating resources was produced by water electrolysis at the service stations.3,4 Hydrogen service stations were assumed to have onsite compression, storage, and dispensing for 5000psi delivery to the vehicle. The potential hydrogen generation from each of the four resources were determined on a state-by-state basis. The U.S. totals for hydrogen potential in 2040 are listed in Figure 1. The hydrogen potential includes the effect of conversion efficiencies (gasifier or electrolyzer) and line losses in transmission. The hydrogen demand for a given state in the year 2040 was estimated by assuming that the per capita demand for hydrogen would be proportional to the per capita gasoline usage in that state.5 To calculate the gasoline need for a state in 2040, we multiplied the predicted population by the current per capita gasoline usage, assuming that the relative gasoline use per capita remains the same for each state in the future. The 10 quads of hydrogen was then divided among the states in the same proportion as the projected gasoline consumption. 1
One quad = 1015 Btu. In this report, the lower heating value (LHV) is used when referring to the energy content of a fuel. 2 Based on an average 2.2X efficiency gain for FCVs over conventional internal combustion engines. 3 Alternate hydrogen production methods were considered, but the lowest cost routes to hydrogen were gasification/reforming from biomass and electrolysis from wind, solar, and geothermal electricity. 4 As an exception to these methods, landfill gas was assumed to be delivered to the service stations via natural gas pipelines and reformed onsite. 5 The per capita gasoline consumption varies considerably from state to state.
1
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
The cost for hydrogen from each resource was calculated by applying discounted cash flow (DCF) analysis for each step along the hydrogen pathway from production to dispensing into vehicles. The capital costs for equipment were taken from the literature or from projections for future costs for renewable technologies. The factors used to calculate the cost of hydrogen are listed in Figure 2. Hydrogen Energy Potential in 2040 quads/year Biomass
2.76
Wind
22.9
Solar
5.9
Geothermal
0.4
U.S. Total
31.9
Figure 1. Hydrogen potential from the major renewable resources in 2040.
Resource Wind Class 4 Wind Class 5 Wind Class 6 PV solar Geothermal Biomass Gasification Electricity transmission (high-voltage) Interstate H2 pipeline Local H2 pipeline Electrolyzer Compression, storage, and dispensing
Capital Costa $648/kWp $648/kWp $648/kWp $1,000/kWp $1,500/kWp $71,091/m.t. H2 per day ($102,000 + $500/MW) per mile $2.79/(mile-106 Btu/day) N/Ac
Capacity Factor 38.3% 41.4% 48.7% 25% 95%
Operating Costsb $0.019/kWh $0.018/kWh $0.016/kWh $0.032/kWh $0.012/kWh
85%
$0.83-$1.26/kg H2d
70%
$0.0016/kWh
60%
$0.064/106 Btu
N/Ac
$2.99/106 Btu
$300/kWth
69%
$0.87/kg H2
$513/(kg H2/day)
69%
$0.27/kg H2
a
Capital costs are in 2001 dollars. Includes operation & maintenance, utilities, and overhead expenses. c Local H2 pipeline cost per unit energy was assumed to be the same as the current cost for local natural gas pipeline. d Biomass operating cost includes feedstock costs ranging from $22/metric ton (for manure and MSW) to $44/metric ton (for dedicated energy crops). b
Figure 2. Factors used to calculate the cost of hydrogen from renewable resources.
6
The H2 potential from biomass listed here accounts for competition for biomass resources for alternate uses, with a maximum of 4.3 quads. See Section 2.1.6 and Appendix A of the full report for details.
2
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
The resulting costs of hydrogen at the dispensing site from the various pathways are shown in Figure 3. The hydrogen costs listed on Figure 3 include the cost for 500 miles of transmission for comparison purposes, although the distribution system described below accounts for actual transmission miles.
7
a
Electricity Electrolysis Comp.,Stor.,Disp. Transmission (per 500 miles)
3.5 3
5
H2 cost, $/kg
H2 cost, $/kg
6
4
2 1.5
2
1
1
0.5
Wind Class 4
Wind Class 5
Wind Class 6
PV Solar
GeoThermal
Feedstock Gas. & Ref. Local H2 pipeline Local NG pipeline Comp.,Stor.,Disp. Pipeline (per 500 miles)
2.5
3
0
b
0
Dedicated Agricultural Wood Energy Residues Waste Crops Residues
MSW
Livestock Manure
Landfill Gas
Figure 3. Cost of hydrogen from renewable resources, including 500 miles transmission from source to vehicle dispensing for (a) electrolysis and (b) reformation production methods.
A simulation was created to generate a hydrogen production scenario for the continental U.S. based on resource availability and cost. The estimated costs of hydrogen are based on generation and transportation costs, with interstate distances taken into consideration. The model strives to simulate an unregulated hydrogen market in which consumers in each state buy hydrogen generated from the least expensive available resource from any other state. The algorithm takes as input the resource availability, base costs, and needs in each state and outputs the resulting distribution of hydrogen resource usage and prices. While the simulation does not deliver the true cost-optimized renewable hydrogen market, it does model a realistic low-cost hydrogen supply infrastructure. In the simulation, hydrogen is purchased in small units (0.0001 quads), with each state purchasing the cheapest available hydrogen in each buying round until all of that state’s hydrogen needs are met or the resources are consumed. At the beginning of each round, a cost for hydrogen generated using each resource from each state to each state is calculated. If a particular resource in a state is non-existent or has been consumed, that resource is unavailable for purchase. The simulation generates a round-by-round history of the hydrogen purchase choices for all states, providing potential resource preferences, locations, and transmission. In order to evaluate the simulation results, we can look at the statistics for hydrogen production on a national basis as well as examining the results for individual states. Figure 4 shows the simulation results for the annual resource availability and consumption. Consumption of each resource depends both on its availability and on the
3
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
cost of hydrogen from that resource relative to all other resources. While biomass and geothermal resources represent a small hydrogen potential relative to solar and wind energy, these resources make significant contributions to the national hydrogen demand due to their relatively low cost compared to other renewable resources. Solar energy, however, makes no contribution since it is considerably more costly than the abundant Class 4 wind energy with which it must compete. Of particular note is that 3.34 quads of the 4.8 quads of potential hydrogen from wind classes 5 and 6 goes untapped. Much of this unused wind potential is in the Midwest and Rocky Mountains, where the high transmission cost to the population centers on the coasts prohibits higher-class wind from being cost competitive with Class 4 wind resources at shorter transmission distances. Resource
Potential (quads/yr)
Usage (quads/yr)
Wind Class 4
18.1
5.3
Wind Class 5
3.1
0.48
Wind Class 6
1.7
0.98
Photovoltaic Solar
5.9
0.0
Geothermal
0.43
0.43
Dedicated Energy
0.57
0.57
Agricultural Residue
0.81
0.81
Wood Waste
0.73
0.73
Municipal Solid Waste
0.18
0.18
Livestock Manure
0.29
0.29
Landfill Gas
0.13
0.13
Figure 4. Summary of hydrogen availability and usage for each resource predicted by the model.
The resulting average cost of hydrogen in each state is listed in Figure 5. Generally, the least expensive hydrogen initially available to all states is from in-state biomass resources, followed by the biomass in neighboring states. Nearly half of all hydrogen from biomass is consumed in-state, with the amount transported out of state (via pipeline) decreasing with increasing inter-state distance. The average inter-state pipeline distance is 259 miles. As states consume their surrounding biomass resources, they are forced to purchase electricity for hydrogen generation, which may be transmitted over longer distances. Only 24% of renewable electricity is used in-state for hydrogen production, and no obvious trend exists for quantity of hydrogen transmitted as a function of transmission distance. The mean distance for renewable electricity transmission is 540 miles. It is interesting to note that no resource is transmitted (via electrical lines or pipelines) more than 1,500 miles, indicating that coastal states reach no further than the middle of the country to meet their hydrogen demand.
4
Hydrogen from Renewable Energy Sources
AL AR AZ CA CO CT DE FL GA IA ID IL IN KS KY LA
$3.92 $3.72 $3.49 $4.09 $3.51 $3.25 $2.60 $4.72 $4.49 $3.11 $2.73 $4.16 $4.01 $3.23 $3.71 $3.63
Directed Technologies Inc.
MA MD ME MI MN MO MS MT NC ND NE NH NJ NM NV NY
$4.17 $4.11 $2.73 $4.17 $3.60 $3.84 $3.48 $2.60 $4.50 $2.54 $2.62 $2.72 $4.45 $3.08 $3.03 $4.52
OH OK OR PA RI SC SD TN TX UT VA VT WA WI WV WY
$4.30 $3.36 $3.33 $4.39 $2.62 $3.98 $2.56 $4.11 $4.03 $3.13 $4.34 $2.57 $3.68 $3.70 $2.64 $2.59
Figure 5. Average cost per kilogram of hydrogen (delivered to vehicle at 5000psi) for the 48 contiguous states. The U.S. average cost is $3.98 per kg of H2.
Conclusions The annual generation of 10 quads of hydrogen in the years 2030-2050 from renewable sources for transportation uses in the U.S. is technically achievable and, according to our model, leads to a national average hydrogen cost of $3.98/kg ($33.24/GJ, LHV basis) delivered to the vehicle. Wind and biomass are the most significant resources (on an energy supplied basis) for hydrogen production, with geothermal playing a small role due to its limited potential. Hydrogen from renewable electricity is expensive compared to that from the reformation of biomass for three important reasons. First, most renewable electricity, especially wind and solar, is expensive compared to electricity from fossil fuels due primarily to high capital costs for the wind and solar installations with relatively low capacity factors (all 150°C, geothermal source to be economically viable. However, there is a significant resource in the U.S. of low-quality geothermal energy that may be economical for direct uses such as space heating, agriculture, and aquaculture. Such application could displace the use of fossil fuels used for heating (direct or through electricity generation), freeing those fuels for hydrogen production in processes such as steam methane reforming. Since this report focuses on hydrogen production from renewable resources, only geothermal sources appropriate for electricity production will be considered. There are three methods of geothermal electricity generation, listed in order of generally decreasing reservoir temperature and correspondingly lower theoretical efficiencies: •
32
Direct steam: Naturally-occurring steam from a geothermal well is used to drive a turbine. Geothermal fluids typically contain impurities, including salts and sulfur compounds, that must be removed or treated before disposing of the condensate.
“Generating Power with the World’s Tallest Tower.” Business Week, 9 September 2002, p. 69.
14
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
•
Flash steam: Similar to direct steam except that the geothermal fluid is high temperature water or a mixture of water and steam that must be “flashed” (exposed to a pressure drop) to separate the usable steam from the water.
•
Binary: The hot geothermal resource (steam, hot water, or two-phase mixture) is used to vaporize a working fluid that has a boiling point lower than water. Isobutane, pentane, and hydrofluorocarbon (HFC) refrigerant fluids are common working fluids in binary plants.33
The direct steam process is the most economical method of the three since steam from the geothermal reservoir can be used with little pretreatment. Flash steam and binary are less economical since more processing steps, and consequently additional equipment, are required for heat recovery from the reservoir. Commercial installations of geothermal electricity plants are generally small compared to utility power plants, with the exception of The Geysers plant constructed in California in the 1960’s. The Geysers nominal capacity is 2,850 MW, or the equivalent of four large coal- or natural gas-fired power plants. The actual power output from The Geysers has declined over the years due to depletion of the steam reservoir. The trend in geothermal electricity generation is toward small distributed plants and away from power-plant level production. 1.2.4 Biomass “Biomass” is generally considered to be any material that is derived from plant or animal matter so that the use of the material for energy generation, which by its nature causes emissions of carbon dioxide, can be offset by the growth of new plant or animal matter to reabsorb a quantity of carbon dioxide equal to that emitted. For the purposes of this study, we have considered the following categories of materials to be biomass:
•
•
•
Energy crops: A plant that is raised specifically to use for energy generation is considered an energy crop, and theoretically any plant could be used for energy or hydrogen generation. The plants most often associated with energy production are switchgrass, hybrid poplar, and willow. Agricultural residues: Stalks that remain in the field after harvest are classified as agricultural residues. Sustainable farming practices require that some crop residue be left in the field to supply organic material to the soil necessary for the next planting season. There may be some residues from food processing (e.g., peanut shells) that could be converted to hydrogen, but these quantities are likely to be small in comparison to harvest residues. Wood waste: Material left in the forest after timber harvesting and the discard such as sawdust from processing are considered wood waste. Much of the waste in saw mills and paper making facilities is already burned to produce steam or
33
Rafferty, Kevin. Geothermal Power Generation: A Primer on Low-Temperature, Small-Scale Applications. Report from the Geo-Heat Center, January 2000.
15
Hydrogen from Renewable Energy Sources
•
•
Directed Technologies Inc.
process heat for the facility, so the availability of this biomass will depend on the ability to find substitute sources of steam and process heat. Municipal solid waste (MSW) and landfill gas: It can be argued that MSW and landfill gas are not renewable in the same way that solar or wind energy is renewable, however, we have included both in this analysis since in most cases the waste would be unused for any other purpose. Livestock manure: High population-density livestock facilities such as dairy farms, hog farms, and chicken houses handle and process large quantities of manure. There is increasing pressure on farmers in some locations to reduce odors from and discharge of livestock wastes to waterways. In many locations the manure must be collected and processed before disposal or land application as fertilizer, so the infrastructure may already be in place to deliver the waste to a central hydrogen production facility.
In the case of biomass resources, three potential pathways to hydrogen exist: • •
•
The first pathway is biomass gasification combined with a gas turbine or with a conventional steam-cycle to generate electricity, in which case the electricity is treated in the same way as solar, wind, or geothermal energy. The second pathway is gasification and steam reformation, generating hydrogen directly from the biomass. Distribution of hydrogen then generally requires a hydrogen pipeline network, possibly supplemented by liquid hydrogen transportation by truck for locations not served by the hydrogen pipeline. The third pathway is conversion of the biomass to liquid hydrocarbons, or biocrude, near the feedstock source, with the biocrude transported by truck or pipeline to biocrude steam reformers at the hydrogen dispensing sites.
The one exception to these cases is reformation of landfill gas. Since landfill gas is predominantly methane and carbon dioxide, the methane can be purified and routed directly into the natural gas pipeline. An equivalent volume of natural gas can be removed from the pipeline at the refueling site, where it can be steam reformed to produce hydrogen. The biomass-to-H2 pathway chosen for a particular feedstock and location will depend on the economics of the distribution system. The pathway for a particular location will be dependent on the availability and cost of distribution infrastructure, and these issues are addressed in detail in Sections 2 and 3.
16
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2 Hydrogen Supply, Demand, and Economics 2.1 Renewable Energy Resources (Supply) The first task in developing a 10 quad/year hydrogen-from-renewables system is to determine that at least 10 quads of H2 are available. This section provides the estimates for total potential hydrogen from each of the renewable energy sources discussed in Section 1.2. Generally, the literature on renewable energy list sources by state, and we aggregated the state numbers to develop the national totals. We have chosen 2040 to be the benchmark year for calculating renewable resource availability (midpoint of the 2030-2050 period), since some resources increase or decrease every year. The potentials of the major renewable resources in this analysis do not depend on the year, as the availability of the resources is only marginally influenced by the state of technology. The quantity of feedstocks such as municipal solid waste, however, are directly proportional to the population and will change every year.
The estimates presented in this section are single values for each of the resources, although there is significant uncertainty in projections for 30 to 50 years in the future. In Section 3 we examine the sensitivity of our predictions to the important assumptions about cost and availability. 2.1.1 Competition for Renewable Energy Resources The use of renewable resources for applications like electricity generation may infringe on the potential resource available for hydrogen generation. Federal and state governments are actively developing or considering requirements for renewable energy portfolios in electricity generation. The U.S. Senate recently approved a provision in the energy bill that would require investor owned utilities to produce a minimum of 10% of their electricity from renewables by 2020.34 With a projected electricity generation of 16.8 quads,35 of which 79.4% would be supplied by investor owned utilities,36 the potential need for renewable electricity to support the regulation is 1.3 quads in 2020. The quantity of electricity needed to meet the 10% requirement may be as high as 2 –3 quads in 2030-2050 with modest growth in electricity usage and an increase in renewables contribution to the electricity supply. 37 2.1.2 Electricity Grid Considerations The major sources of renewable energy, with the possible exception of biomass, directly generate electricity, with the electricity converted to hydrogen by electrolysis of water. Since electricity is prohibitively expensive to store in quantities on the scale of this analysis, the electricity must be used when it is generated. In this study we trace the “electrons” from the source through the transmission systems to hydrogen only to calculate the cost of hydrogen from a particular renewable electricity source.
34
U.S. Senate, “Energy Policy Act of 2002,” April 25, 2002. Based on H.R. 4. Department of Energy, Energy Information Administration, Annual Energy Outlook 2002. 36 Department of Energy, Energy Information Administration, Annual Energy Review 2000. August 2001. 37 The U.S. DOE in the Annual Energy Outlook 2003 projects renewable electricity generation (excluding hydro) to be 131 billion kWh (~0.45 quads) in 2025 in the absence of new federal mandates. 35
17
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
To simplify our estimates, we have ignored two important realities of renewable electricity production: 1) production intermittency for wind and solar resources as it affects grid capacity and electricity demand; 2) peak versus off-peak electricity pricing. Ignoring the first issue means that hydrogen prices reflect the average cost of electricity production at the renewable resource site and the cost of transmitting electricity for electrolysis to the service station, but not the cost of extra grid capacity to account for the sporadic electricity production. Ignoring the second means that while the actual electricity produced from the resources can be used by any electricity-consumer on the grid, the cost of renewable electricity is paid for by hydrogen consumers alone. In reality, hydrogen production will be timed to coincide with periods of low electricity prices, assuming that there is enough buffer in the hydrogen storage system to outlast the periods of high electricity prices. For example, solar electricity is generated only during daylight hours and could be sold to electricity users at a favorable price during peak-use hours, but hydrogen production may be scheduled at night when the cost of electricity from the grid is lower. We have also assumed that the electricity grid is stable, that is, it provides a consistent voltage and load that meets the demand, no matter what the mix of renewable sources of hydrogen. Wind and solar are both intermittent resources that can not be reliably predicted on a continuous basis, so there is a maximum contribution from renewables that can be compensated for by the non-renewable electricity production methods. Denmark, a country with huge offshore wind electricity installations, has experienced grid instability at about 20% contribution from wind.38 The efficiency of electrolyzers is defined as the energy content of the hydrogen divided by the electrical energy input. The projected efficiency based on the hydrogen lower heating value (LHV) of future electrolysis processes is anticipated to be 76%.39 At this projected efficiency, 1.32 units of electricity are required to generate each unit of hydrogen energy in the electrolyzer. Including typical electricity transmission line losses of 6% system-wide,40 1.4 units of electricity must be generated at the source to produce one unit of hydrogen energy at the electrolyzer. 2.1.3 Wind Our estimate for hydrogen recoverable from wind energy is based on the following assumptions:
•
All U.S. land of Classes 4, 5, and 6 wind in the contiguous 48 states and within 10 miles of 100 kV or greater transmission lines is utilized, with the following exclusions: parks, urban areas, water, 50% of forest land, 30% of agricultural land, and 10% of range land.41 This assumes all qualifying land is put to use,
38
Insert the Denmark reference. Ogden, J.M. and J. Nitsch. “Solar Hydrogen,” in Renewable Energy: Sources for Fuels and Electricity. Johansson, Kelly, Reddy, and Williams, eds. Island Press, Washington, D.C., 1993. 40 U.S. Department of Energy, Energy Information Administration, Annual Energy Review 2000. 41 EPRI and U.S. Department of Energy. Renewable Energy Technology Characterizations, TR-109496, December 1997. 39
18
Hydrogen from Renewable Energy Sources
• • • •
Directed Technologies Inc.
although a maximum of only 5% of the land is actually covered by the wind turbines and associated equipment. The remaining 95% of the land area can still be dedicated to other uses, for example, farming, livestock grazing, or even solar electricity generation. The contribution of each wind class to an individual state’s total wind resource was determined from the wind resource map (Figure 1-2). Capacity factors42 are 38.3% for Class 4, 41.4% for Class 5 and 48.7% for Class 6 based on projections for improved technology.43 Availability factor44 is 98%. Turbines are assumed to be large-scale (1 MW each), with a 50-meter minimum hub height, the standard for larger turbines.
After accounting for an electrolyzer efficiency and transmission losses, the quantity of hydrogen that could be produced from wind energy is • • • •
Class 4: Class 5: Class 6: Total:
18.1 quads 3.1 quads 1.7 quads 22.9 quads
Electricity generation from wind must grow at a 15-16% compound annual rate to increase from the 2000 level of less than 0.1 quad45 to the 2040 level of 22.9 quads. Sustaining such a high growth rate in wind power installations may be difficult, since the highest quality, and consequently lowest cost, wind sites will be developed first, leaving the lower quality sites undeveloped. However, since we are developing a pathway to only 10 quads of H2, a maximum of 45% of the wind resources will be tapped even if no other renewable energy sources are included. There are substantial sources of Classes 5-7 wind off the Oregon and Massachusetts coasts, which could yield powerfully efficient energy pending public aesthetic acceptance, advancements in offshore turbine technology, and ease of access. The offshore wind resources have not been included in the totals listed above due to the low probability of significant offshore wind development by 2030. The outcome of current and near-term offshore wind projects throughout Northern Europe will likely determine if the U.S. will follow suit and utilize these resources. The projections used to estimate the wind potential did not assume any extension of the electricity grid to reach additional remote locations. It is likely that by 2030-2050 there will be more Classes 5 and 6 land areas that meet the 10 miles-to-transmission line qualification. Although the total Classes 5 and 6 potential would not reach the levels of
42
“Capacity factor” is the ratio of the average rate of power production to the theoretical capacity. Ref. 41. 44 “Availability factor” is the fraction of time the generator is operational, considering maintenance downtime. 45 Energy Information Administration. Annual Energy Review 2000. 43
19
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Class 4, the lower costs of Classes 5 and 6 electricity compared to Class 4 (see Section 2.4) may influence the decision to use wind instead of other resources. Furthermore, the capacity factors for these wind classes are expected to improve with better turbine engineering and advances in rotor materials and design. Capacity factors in this study were taken from 2001 projections for 2030, although there have already been demonstrations of 48-52% capacity factors.46 2.1.4 Solar The only practical restrictions on the availability of solar resources are the cost (to be considered in Sections 2 and 3) and the land area required. The favorable locations for using PV solar electricity to produce hydrogen are in the desert Southwest (e.g., Arizona, New Mexico, southern California) since other southern locations (e.g., Florida, Texas) have more cloud cover and an insolation approximately 10-15% lower than in the desert region. On a national level, only the most efficient solar resources will be exploited for utility-scale installations that are required to make an impact on the 10-quad scale. Rooftop solar panels on commercial buildings and private residences are more likely to generate electricity for local consumption (offsetting consumption of electricity from power plants) than to provide electricity to the grid for hydrogen production.
The Bureau of Land Management (BLM) controls 262 million acres in the western U.S., with approximately 22.3 million acres (8.5%) of that land in prime solar areas that have average annual insolation > 6 kWhr/m2-day. The land area required to generate one quad of hydrogen in these regions is projected to be about 378,000 acres.47 If 10% of the BLM land is used for solar electricity generation, the quantity of hydrogen that could be generated is 5.9 quads. It is uncertain whether or not nearly 1% of BLM land will be developed for solar electricity generation, however, this projection provides a reasonable range for solar potential. As we demonstrate in Section 3, unless the cost of solar electricity can be reduced below the level projected for 2040, there will be few, if any, utility-scale solar electricity generation installations and land availability will not be an issue.48 2.1.5 Geothermal The best estimate of the geothermal potential for electricity generation in the U.S. with optimistic assumptions about improvements in drilling and reservoir management
46
“The Potential of Wind, Solar, Geothermal for the West”, Dr. A. Leitner, RDI Consulting, NEMS/Annual Energy Outlook 2002 Conference. 47 Assumes a 25% solar-to-electric conversion efficiency (an optimistic projection for future technology), 50% land coverage for solar modules and balance of system equipment, 76% electrolyzer efficiency (LHV basis), and 6% electrical transmission line losses. 48 In 2030-2050, solar electricity may be developed enough to play a significant role in distributed power generation or summer peaking, since the cost of solar electricity may be competitive with peak prices in some markets. However, solar electricity is unlikely to be cost-competitive with wind when there is a choice between the two.
20
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
technologies is 18,800 MW.49 We assume that the geothermal electric capacity has been realized by 2030 and remains essentially flat through 2050. The hydrogen yield from electrolysis of 18,800 MWe operated 24 hours/day, 365 days/year is 0.43 quads. The lack of cooling water in desert locations, especially in Nevada and Southern California where geothermal resources are most prevalent, may limit the number of locations that can be exploited for geothermal electricity generation since cooling tower water is a much preferred heat rejection medium than is air. The cooling water can theoretically be cooled to the ambient wet bulb temperature, which is often substantially below the air temperature in arid climates. The theoretical maximum energy efficiency calculated using Equation 1-6 increases approximately 2.4% for every 10°C reduction in cooling fluid temperature.50 2.1.6 Biomass In the years 2030-2050 there are likely to be many competing uses for biomass resources and land, including electricity generation by co-firing biomass with fossil fuels, conversion to liquid fuels such as ethanol and biodiesel, and increased demand for food as the population increases. Accurately projecting the demand for any biomass use 30 to 50 years in advance is difficult, so a logical and defensible allocation of biomass resources must be estimated. The biomass resource allocation method used in the subsections to follow is based on limiting availability by capping the biomass feedstock costs. Capping the cost of each biomass resource for the production of H2 for transportation uses frees the remaining biomass for other end uses that are not a concern of this analysis.
To calculate the quantity of hydrogen that could be generated from biomass, we have assumed that the biomass is gasified and steam reformed to produce pure hydrogen. This pathway produces the highest hydrogen yield per unit of biomass, so any other process that produces hydrogen from biomass will have a lower hydrogen potential. Based on simulations of the gasification process, approximately 70 kg H2 (net) can be produced per dry metric ton of biomass. This yield is consistent with literature values from commercial gasification processes51 and is constrained primarily by thermodynamic limitations that are unlikely to be overcome by improvements in technology in future years.52
49
Gawell, K, et al. Preliminary Report: Geothermal Energy, the Potential for Clean Power from the Earth. Geothermal Energy Association, April 1999. 50 Calculated for a 150°C reservoir temperature with 35°C, 25°C and 15°C cooling fluid temperatures. The resulting efficiencies were 27.2%, 29.6%, and 31.9%, respectively. 51 Craig, K.R. and M.K. Mann. Cost and Performance Analysis of Biomass-Based Integrated Gasification Combined-Cycle (BIGCC) Power Systems, National Renewable Energy Lab Report, October 1996. 52 There is a possibility that integrated gasifier/separation processes-for example, membrane reactors-that remove H2 from the reaction zone to force the equilibrium toward higher H2 production will be developed by 2030. However, we do not anticipate that membrane reactors will contribute a significant fraction of the for the large-scale installations required for widespread H2 production.
21
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.1.6.1 Dedicated Energy Crops
De La Torre Ugarte, et al. conducted a modeling study to determine the potential for conversion of existing agricultural land to bioenergy cropland by the year 2008.53 The model calculates the effects of competition between energy crops and existing crops, livestock grazing, and Conservation Resource Program (CRP) land. The land area that would be devoted to energy crops in the model scenario is 41.87 million acres with a total energy crop of 188.1 million dry tons for a yield of 4.5 dry tons per acre at a price of $50/dry ton delivered. In this case, the only energy crop expected to be economically competitive is switchgrass, with all production east of the Rocky Mountains. The average switchgrass yield falls within the range of similar traditional crops—in 2000 the average yields were 2.54 tons per acre for hay and 16.8 tons per acre for corn for silage.54 We assumed that there are no changes in crop distribution or total acreage between 2008 and 2030 and that the productivity of energy crops increases at the same rate as traditional crops.55 The energy crop harvest in 2030 is estimated to be 212.0 million dry tons in 2030 and 236.3 million dry tons in 2050. If the entire annual energy crop at the $50/ton price were converted to hydrogen, the resulting hydrogen energy would be 1.51.7 quads. We have assumed that the biomass actually available will be less than the potential maximum. Walsh, et al. estimated the size of the energy crop that would be available at a range of prices up to the maximum of $50/ton. 56 We have used the energy crop size at the $40/ton level, 35.2% of the maximum, for this study. We believe that the 35.2% availability reasonably represents the energy crop potential for hydrogen production when competing uses for the energy crop are considered, especially other power generation options such as co-firing with coal. Thus, the hydrogen available from energy crops in the years 2030 to 2050 ranges from 0.54 to 0.60 quads. The cost for biomass including 2.1.6.2 Agricultural Residues and Wood Waste
The quantities of agricultural residues and wood waste available for energy use at a range of price levels were estimated from the same modeling study by Walsh, et al. related to the dedicated energy crop analysis.57 Since some fraction of the residues must be left in the field to maintain the soil quality, the model assumed that only 30-40% of the residues could be removed from the field. The model considered only corn and wheat residues since other major crops such as soybeans provide very little residue and the harvest size of most other crops is small in 53
De La Torre Ugarte, D.G., et al. The Economic Impacts of Bioenergy Crop Production on U.S. Agriculture. July 2000, available at http://bioenergy.ornl.gov/pubs/econ_assess.html 54 USDA National Agricultural Statistics Service historical data, http://www.usda.gov/nass 55 The yield improvement for traditional crops was calculated from the USDA annual summary of yield for hay and corn for silage, which should be a similar type of farming to switchgrass. Since 1980 the average annual rate of yield improvement for those two crops was 0.55%. 56 Walsh, M.E., et al. Biomass Feedstock Availability in the United States: 1999 State Level Analysis. January 2000, available at http://bioenergy.ornl.gov/pubs/econ_assess.html 57 Ref. 56.
22
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
comparison to corn or wheat. The potential agricultural residue available at the highest price considered ($50/ton) is 151 million dry tons, with a hydrogen potential of 1.09 quads. Wood waste includes cuttings left in the forest after timber harvest, mill waste from lumber processing, and urban wood waste (construction and tree trimming). Currently, some wood waste is burned to provide steam and heat in the processing facility, so the processor must switch to a replacement energy source. The potential wood waste available at the highest price considered ($50/ton) is 135 million dry tons, with a hydrogen potential of 0.98 quads. Agricultural and wood wastes are not available at zero cost because there is significant cost incurred to collect, prepare, and deliver the materials to the facility where the conversion to hydrogen takes place. As a result, only 74% of the potential agricultural and wood wastes would be available at a cost of $36/ton, reducing the hydrogen available to 0.81 and 0.72 quads, respectively. 2.1.6.3 Municipal Solid Waste
In 1998, the average person in the U.S. generated 3.2 pounds per day of MSW, net of recycle.58 If this rate remains steady through the period 2030 to 2050, the total trash generated will increase linearly with the population. We have assumed that the hydrogen yield from MSW will be 58 kg/dry metric ton instead of 70 kg/dry metric ton because MSW includes a significant fraction of glass and metal that contains no recoverable hydrogen.59 The quantity of trash that would be generated in 2030 with population of 351.1 million is 205 million tons and in 2050 with population of 403.7 million is 236 million tons. 60 The hydrogen energy recoverable from these quantities of MSW is 1.22 to 1.40 quads. From the model of Walsh, et al., the projected MSW that would be recovered at a cost of $20/ton is 14.0% of the total in 2030 to 2050, or 29-33 million tons/year. The hydrogen yield from the reduced quantity of MSW is 0.17-0.20 quads. Our projections represent an upper limit for hydrogen from MSW, since a reduction in the per capita generation of trash is likely as landfill space is depleted and alternative uses for MSW such as electricity generation may be favored economically. 2.1.6.4 Landfill Gas
Municipal solid waste landfills generate gas due to anaerobic decomposition of organic matter in the trash. The landfill gas (LFG) contains on average about 50% methane that can be converted to hydrogen by steam reforming. The number and size of municipal landfills and the amount of LFG collected or flared per ton of waste in place (WIP) were
58
EPA Environmental Fact Sheet. Municipal Solid Waste Generation, Recycling and Disposal in the United States: Facts and Figures for 1998. April 2000, www.epa.gov/osw 59 Klass, D.L. Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San Diego, 1998, page 140. 60 A detailed explanation of the how the population projections are calculated is in Section 2.2.2
23
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
estimated from the EPA database for 1995.61 It is estimated that in 1995 landfills contained 5.8 billion tons of waste with potential gas generation of 158 scf/year/ton. If the current rate of waste disposal (3.2 lb/person/day) continues until 2030, the total amount of WIP will increase to 11.9 billion tons. For the purposes of our calculations, no new additions will be made to landfills starting in 2030 since all MSW will be converted to hydrogen directly (see Section 2.1.6.3). The total landfill gas generation is projected to be 953 billion scf/year. We have assumed that only 20% of the landfill gas, or 191 billion scf/year, will be recovered for hydrogen production, since currently some landfill gas is recovered for fuel or electricity generation and this practice is expected to become more widespread. The hydrogen potential from the steam reforming of 191 billion scf/year of methane is 0.13 quads. In practice, the calculated rate of landfill gas production can only be maintained for a limited number of years, since no additional MSW is accumulated, and in later years an increase in another renewable resource will be required to compensate. We have not accounted for the decline of landfill gas production since the contribution to the total quantity of hydrogen is relatively small, however, any focused analysis of landfill gas should address this issue. 2.1.6.5 Livestock Manure
The annual quantity of manure available nationwide was estimated by multiplying a “manure generation factor” for each type of livestock62 by the population of livestock in each state.63 We included only cattle, hogs, chickens, and turkeys since the populations of all other types of livestock are negligible in comparison. It is unlikely that all livestock manure could be recovered for hydrogen conversion. Like other renewable resources, this is more an issue of practicality and economics than technology. Only the manure generated in barns or other high-density locations would be available for capture and conversion to hydrogen. Klass64 includes a survey of potential for recovery of livestock manure as an energy source. The “collectible” fraction of manure ranges from 6.5% for beef cattle to 85% for hogs, with the other animals between the extremes. We have assumed that 20% of the manure generated annually will be collected for conversion to hydrogen, since it is reasonable to assume that much of the manure will continue to be used as fertilizer. The manure generation for each type of animal included is listed in Figure 2-1. A breakdown of manure generation by state is provided in Appendix B.
61
U.S. EPA database of municipal landfills, http://www.epa.gov/epaoswer/nonhw/muncpl/landfill/index.htm#list 62 Barker, J.C., et al. North Carolina Agricultural Chemicals Manual, 2002. 63 U.S. Dept. of Agriculture, livestock census 1997. 64 Klass (ref. 59), pp. 145-146.
24
Hydrogen from Renewable Energy Sources
Animal Type
Manure wet tons/yr/animal
Heifer Dairy cattle Beef cattle Hogs Chickens Turkeys
Directed Technologies Inc.
Population (thousands)
22.3 27,000 22.3 9,100 8.3 34,100 1.9 61,200 0.0355 1,464,500 0.112 98,950 Total Total @ 20% recovery for H2 production
Manure 10-6 × wet tons/year 602.1 202.9 283.0 116.3 52.0 11.1 1,267 253.5
Hydrogen Potential quads/year 0.70 0.24 0.33 0.13 0.06 0.01 1.47 0.29
Figure 2-1. Livestock manure generation by animal type. Wet manure is approximately 14.5% dry matter, and hydrogen yield is 7% by weight on a dry basis.
A separate estimate for the quantity of livestock manure generated annually in the U.S. from cattle, hogs, chickens, and turkeys is 1.36 billion tons (wet basis).65 Wet manure typically has a solids content of ~15%, so the dry matter available from the manure would be 204 million tons, with a theoretical hydrogen recovery of 1.26 quads. The close agreement between this number and the totals in Figure 2-1 give credibility to the stateby-state analysis. 2.1.6.6 Biomass Summary
As discussed in the introduction to Section 2.1.6 and quantified in the dedicated energy crop and agricultural/wood waste sections, competing uses for biomass feedstocks will reduce the quantity of biomass available for H2 generation. The estimated H2 from biomass of all types for transportation uses is 2.7 quads, or 63%, out of a potential total of 4.3 quads.66 The Vision for Bioenergy & Biobased Products in the United States prepared for the U.S. Congress sets goals for quantities of or relatively market share for biomass-derived products in the years through 2030.67 Although the goals set in the Vision are aggressive and exceed the biomass potential used in this analysis, the split among biopower, biobased transportation fuels, and biobased chemicals/materials is useful for comparison to the 63% for transportation uses estimated above. The Vision goal is the use of approximately 15.9 quads of biomass in 2030, of which 9.4 quads (59%) are projected for use as transportation fuels.68 The Vision fraction of biomass for transportation fuels is in close agreement to the 63% assumed in this report. A scenario in which the total biomass availability is used for H2 production was considered as an alternate case, and the detailed results for the alternate case are presented in Appendix A.
65
Doyle, Michael. Keeping Foodborne Pathogens Down on the Farm. Presentation available at http://www.fsis.usda.gov/OPPDE/animalprod/Presentations/KFPDF%20Aug%2001/ 66 The 4.3 quads of H2 corresponds to 9.6 quads of primary biomass energy. 67 Vision for Bioenergy & Biobased Products in the United States. October 2002. Available at http://www.bioproducts-bioenergy.gov/pdfs/BioVision_03_Web.pdf 68 This estimate of 59% of biomass resources for transportation requires the use of information sources in addition to the Vision report.
25
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.1.7 National Totals The quantity of hydrogen from renewable sources for 2030 and 2050 that we estimate to be technically achievable is listed in Figure 2-2. The state-by-state breakdown of hydrogen production is provided in Appendix B. H2 in 2030 (quads)
H2 in 2050 (quads)
Wind Solar Geothermal Biomass
22.90 5.9 0.43 2.66
22.90 5.9 0.43 2.76
Average H2 for 2040 (quads) 22.90 5.9 0.43 2.71 (4.3)a
Energy crops Agricultural residues Wood waste Municipal solid waste Landfill gas Livestock manure
0.54 0.81 0.72 0.17 0.13 0.29
0.60 0.81 0.72 0.20 0.13 0.29
0.57 0.81 0.72 0.18 0.13 0.29
Total Hydrogen
31.89
31.99
31.94 (33.5)a
a
The higher number reflects the potential if all available biomass were used for H2 rather than split with competing uses. Figure 2-2. Summary of hydrogen potential from renewable sources.
From the totals in Figure 2-2 it is clear that the production of 10 quads of hydrogen for transportation use from renewable energy sources is well within the technical limits of U.S. resources. The important factor in the development of a hydrogen-vehicle economy then becomes the cost of converting the renewable energy to hydrogen and distributing the hydrogen to the end user, topics that will be addressed in Sections 2.3, 2.4, and 3. 2.1.8 Other Sources of Renewable Energy for Hydrogen We have not included in this study any source of hydrogen that has the potential to generate less than 0.1 quads H2/year by 2050. The following renewable sources of H2 do not meet the minimum threshold and were excluded from the total availability listed in Figure 2-2. 2.1.8.1 Sewage Sludge
Conversion of sewage sludge, or biosolids, to energy may be an economically favorable method for using an otherwise wasted stream. However, the potential quantity of hydrogen from biosolids is small. Based on the biosolids generated by publicly owned treatment works (POTW’s) in 1998,69 the maximum hydrogen recoverable is 0.05 quads.
69
Biosolids Generation, Use, and Disposal in the United States. EPA Report EPA530-R-99-009, September 1999.
26
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.1.8.2 Biological Processes
There is a great deal of current research activity focused on producing hydrogen using bacteria or other simple organisms such as algae that are either naturally occurring or genetically modified. The research programs are too new to make even semi-quantitative predictions for hydrogen production potential in 2030. With current specific hydrogen production rates recently reported in the literature,70 the volume of bacterial culture necessary to generate enough hydrogen for a single fuel cell vehicle (~0.5 kg H2/day) is 77,000 liters. Several orders of magnitude improvements in specific yield are necessary before biological production of hydrogen would be developed at large scale. 2.1.9 Other Issues 2.1.9.1 Land Availability
Solar, wind, and biomass energy crops require substantial land areas to produce energy on the scale considered in this analysis. The land requirements are factored into the energy crop distribution and cost reported in Ref. 53 in Section 2.1.6. The Bureau of Land Management (BLM) is currently studying the potential for renewable energy production on federally-managed lands and has drafted a proposal for land use conditions and rental fees for wind energy development. The BLM proposal suggests royalty payments of 2% of revenues as a starting point for wind energy royalty payments. 2.1.9.2 Water Sources and Costs
Both electrolysis and biomass gasification/reforming require relatively large supplies of water. Electrolysis in particular requires ~2500 gallons of water per metric ton of hydrogen, as water impurities may result in contamination of the hydrogen and subsequent problems with fuel cell operation. We assume that the hydrogen generation processes will include a deionization step for final purification of potable water. Generating 10 quads of hydrogen from water electrolysis, assuming 100% conversion of the water, would require 207 billion gallons of water annually. To put the water needs in perspective, the U.S. potable water use is ~14,750 billion gallons per year, with industrial and agricultural non-potable water use adding ~133,000 billion gallons.71 For the U.S. as a whole, the marginal increase in water use for hydrogen generation would be negligible—only 0.16%. However, in some regions, particularly in California and the Southwest, the water supply is so limited that new sources of fresh water will be required to support the hydrogen infrastructure. The best possibility for expanding the water supply would be desalination of ocean water. Current desalination costs are approximately $2-$5 per 1000 gallons,72 which would add less than 2¢ per kilogram to the cost of hydrogen. This additional cost for water is unlikely to affect the economics of hydrogen production in any location near the coasts with the highest population density where the highest demand for hydrogen is likely to occur. 70
Freemantle, M. “Can We Exploit Hydrogenases?” Chemical & Engineering News, pp. 35-39, 22 July 2002. 71 U.S. EPA website, http://www.epa.gov/safewater/wot/howmuch.html 72 Ettouney, H.M., et al. “Evaluating the Economics of Desalination.” Chemical Engineering Progress, 98 12:32-39 (2002).
27
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.2 Geographic Distribution of Hydrogen Demand In this section we map the likely geographical distribution of hydrogen demand. Ideally, the supply would match demand in each state or county, leading to a trivial distribution plan. However, since supply and demand are generally not well matched geographically, the supply and demand are matched using a mathematical model described in Section 3. 2.2.1 Per Capita Gasoline Consumption We assumed that the demand for hydrogen in each state would be directly proportional to gasoline usage in that state since the 10 quads per year of hydrogen will displace gasoline as the primary passenger vehicle fuel. To estimate the gasoline (i.e., transportation energy) needs in 2030-2050, we assumed that the relative transportation energy need for each state scales with population, as determined by the ratio of population to gasoline consumption shown in Figure 2-3.73 We assumed that the relative gasoline use per capita remains the same for each state in the future. In general, the outliers from the linear relationship between population and gasoline consumption are states with high population concentration in urban areas with low vehicle ownership rates (e.g., New York) and states with relatively higher rural populations with high vehicle ownership rates (e.g., Texas). 2.2.2 State-by-State Population Projections We used the middle series projection from the Census Bureau for the years after the census in 2000.74 The Census Bureau projects state populations for the years 1995, 2005, 2015, and 2025 and national population for every year up to 2100. Growth rates (calculated as % population change over ten years) become fairly level by 2025. The growth rate in each state in the year 2025 was used to calculate the predicted state populations for year 2040. This estimate was then adjustedto the 2040 national population estimates by taking the predicted state population distribution (as % of national total) and multiplying by the 2040 national estimate. The time series (series A) state estimation and mid-level immigration national estimates were used.
73
The gasoline usage was taken from the 1999 DOE data http://eia.doe.gov/pub/state.data/pdf/summaries.pdf and the population from the 2000 Census. 74 U.S. Census Bureau population projections, http://www.census.gov/population/www/projections/natsum.html
28
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2,400 2,200
Gasoline Consumption (10^12 Btu/year)
2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
Population (thousands)
Figure 2-3. Correlation between gasoline usage and population by state (1999 data for gasoline usage, 2000 data for population).75
2.3 Capital Costs One of the benefits of renewable energy sources compared to conventional fossil fuels is that the “feedstock” is available at zero (wind and solar) or relatively low (biomass and geothermal) cost. As a result, the initial capital cost is the primary factor influencing the cost of energy from renewables. We have used current-day capital cost estimates for processes that are well-developed, such as biomass gasification, and future projections for emerging technologies such as wind and solar. The projections for wind, solar, and geothermal capital costs may be optimistic, as there is not a clear path to the projected costs for any of the three, with the possible exception of wind.
The capital costs used in this report are estimates or projections from publicly available sources or widely accepted correlations. We have not developed detailed designs for each of the renewable energy systems since the focus of this analysis was the overall hydrogen production and distribution pathway. We have attempted to avoid using the most optimistic or the most pessimistic scenario for projected costs. In Section 3.3 we examine the sensitivity of the cost of hydrogen to deviations in the capital costs presented here to quantity the effect of uncertain projections on the cost of hydrogen. For consistency, all of the capital costs were converted to end of year 2001 dollars by applying the Chemical Engineering Plant Cost Index (2001 = 394.3).76
75 76
See Ref. 73. Chemical Engineering, August 2002.
29
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.3.1 Wind Current wind turbine and balance of system costs are approximately $800-1,000/kWpeak. Estimates for future reductions in wind turbine costs vary, although the consensus is that reductions in cost will be gradual since wind is the most well-developed of the renewable electricity technologies. The DOE and EPRI in 1997 developed a roadmap for reducing costs for wind electricity generation.77 The 1997 projection for wind energy capital costs for a 50 MWpeak wind farm in 2030 was $635/kWpeak ($648/kWpeak in 2001 dollars). 2.3.2 Solar The current (2002) commercial solar prices are in the range of $5-$7 Wpeak at standard test conditions, depending on the type of cell and tracking system complexity. The NREL long-term goal for PV solar installations, including balance-of-plant items such as transformers, is $1/Wpeak. The progress of cost reduction for recent solar PV installations $16
$14
Installed System Cost ($/Wpeak)
$12
$10
$8
$6
$4
$2
$0 1980
1990
2000
2010
2020
2030
2040
2050
Year
Figure 2-4. The cost of photovoltaic solar systems (including solar module and balance of system components) by year of installation. The individual data points were taken from several reports from actual PV solar installations.78, 79, 80
is plotted in Figure 2-4. By fitting an exponential curve to the data in Figure 2-4, the capital cost for an installed PV solar system is projected to be $1-$1.50/Wpeak in the year 2040, so the NREL goal does not seem unreasonable as a projection for 2030 to 2050. The unknown factor that could dramatically reduce the solar cell cost (but not balance of system costs) is the development of relatively efficient and stable organic solar cells. The 77
EPRI and U.S. Department of Energy. Renewable Energy Technology Characterizations, TR-109496, December 1997. 78 Tomas, M.G., H.N. Post, and R. DeBlasio. Prog. Photovolt. Res. Appl., 7 (1999) 1-19. 79 Solar Electric Power Association. Large Systems Cost Report, 2001 Update, September 2001. 80 Personal communication with Dan Johns, BP Solar, 27 June 2002.
30
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
potential exists for organic solar cells to be manufactured using low-cost techniques such as screen or ink jet printing with inexpensive polymers. However, it is unlikely that organic solar cells will become efficient or stable enough by 2030 to significantly displace the conventional semiconductor cell technology for the widespread installations that will be needed to generate solar electricity on the quad-scale. 2.3.3 Geothermal The costs for geothermal-electric plants cover a wide range, from relatively low cost direct steam plants to relatively high cost binary systems. Most future geothermal facilities will be binary plants because of the relative abundance of low-temperature reservoirs. Meidav, et al. estimate that binary geothermal plants cost $2,500$4,000/kWpeak in 1994.81 We have projected (somewhat optimistically) that the capital cost for geothermal electric plants will decrease to $1,500/kWpeak in 2040 (2001 dollars). Since geothermal will contribute a maximum of 0.4 quads of H2, uncertainties in the capital cost estimate will not have a significant impact on the national distribution and cost of hydrogen. 2.3.4 Biomass to Hydrogen There are several biomass gasification processes that have been commercialized and tested on a pilot or an industrial scale. The technologies may be either oxygen/air gasification, steam gasification, or a combination of the two. The review by Klass provides a detailed breakdown of the subprocess costs for four well-known biomass gasification processes for methanol synthesis with 1650 dry metric tons/day biomass feed rate: Wright-Malta, Battelle, IGT, and Shell Oil.82 The hydrogen production capacity for a 1650 ton/day biomass gasification plant is ~115 metric tons/day, clearly a large-scale industrial gas facility rather than distributed production.
We included in the capital cost estimate only the subprocesses that would be required to produce hydrogen and chose the lowest-cost process, in this case the Battelle steam gasification process operating at atmospheric pressure. The costs for subprocess components for the Battelle process taken from Klass are listed in Figure 2-5. We are not endorsing or dismissing for technical reasons any of the aforementioned gasification processes. The only criterion for the selection is that we expect the cost of gasification processes in 2030-2050 to be at the low end of current designs, and the Battelle process is the lowest cost at the present time. There may be operational differences that favor one of the other processes in certain situations, for example, with a particular biomass feedstock.
81
Meidav, T. and J. Pigott. Geothermal Resources Council Bulletin, 23:10 (1994) 339-344. Klass, D.L. Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San Diego, 1998, p. 323. 82
31
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Process Component
Capital Cost, Thousand $
Feed Preparation
20,500
Gasifier
7,960
Reformer Feed Compressor
12,100
Reformer
17,100
Shift Reactors
10,300
Purification
16,000
Utilities/auxiliaries
33,400
Total
117,300
Figure 2-5. Biomass gasification capital costs for the atmospheric pressure Battelle steam gasification process, adapted from Reference 82. The costs have been adjusted to 2001 dollars.
2.3.5 Biomass-to-Electric The biomass-to-electric plant was assumed to be a 122 MWe low pressure indirectly heated gasifier with an advanced technology utility turbine and steam cycle, which was the lowest-capital cost option in a 1996 NREL study.83 We assumed that the overall plant efficiency will improve from ~35% currently to 45% in the future but that the equipment costs will not significantly change. The costs for the plant subsystems are listed in Figure 2-6, adjusted to 2001 dollars. Process Component
Capital Cost, Thousand $
Biomass feedstock handling and drying
10,859
Gasifier and gas cleanup
22,129
Gas turbine
19,682
Steam cycle
28,718
Balance of plant
56,587
Total
137,974
Figure 2-6. Biomass-to-electricity plant capital costs for a 122 MWe biomass gasification/combined cycle system. Adapted from Reference 83, adjusted to 2001 dollars.
We included only large-scale gasification electricity generation plants since the economies of scale allow for a reasonable cost of electricity. There may be some instances in which it is economically favorable for distributed electricity generation, for 83
Craig, K.R. and M.K. Mann. Cost and Performance Analysis of Biomass-Based Integrated Gasification Combined-Cycle (BIGCC) Power Systems. NREL/TP-430-21657, October 1996.
32
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
example, if the biomass transportation cost is more than we have assumed. However, we believe that the only way for biomass-derived electricity to be competitive with other renewable sources is at the large scale (see Section 3). 2.3.6 Landfill Gas Landfill gas is a special case since the “biomass” feedstock is methane gas. We have assumed that the methane is separated from the methane/carbon dioxide mixture using a conventional amine absorption process, then the purified methane is injected into the local natural gas distribution system. The quantity of methane generated by the landfill is assumed to be reformed to hydrogen at the vehicle fueling station. The methane purification sizing calculations used the median landfill size from the EPA Landfill Database.84 The capital cost for the amine absorption unit was estimated using cost factors for the main equipment (absorber and reboiler) and scaling factors for peripheral equipment, installation, and engineering costs. The methane reforming capital and operating costs were taken from previous estimates by Directed Technologies for distributed reforming of natural gas.85
For electricity generation using landfill gas as the fuel, the median size landfill produces enough gas to support a 5.56 MW turbine operating at 85% capacity factor. The capital cost for a gas turbine in the years after 2020 is projected to be $357 kW (corrected to 2001 dollars).86 The capital cost for the 5.56 MW turbine is projected to be approximately $2 million. 2.3.7 Hydrogen Pipelines There may be a need for either interstate/large capacity, local/small capacity hydrogen pipelines, or both, depending on the renewable energy source and distance from the user. There are hydrogen pipelines in existence now in geographic areas with extensive chemical processing and petroleum refining activities (e.g., Louisiana-Houston and Los Angeles). However, the current hydrogen pipeline network would be insufficient to support a nationwide fleet of fuel cell vehicles.
The natural gas transmission and distribution pipeline network provides a model on which to base the potential hydrogen distribution system. A rule-of-thumb estimate for the capital cost of large-scale hydrogen pipelines is a factor of 1.4-times more, on the basis of energy-transmitted, than equivalent natural gas pipelines.87 The construction of a natural gas pipeline between Wyoming and California, covering a distance of 717 miles, with a capacity of 840 billion Btu/day was recently approved.88 The cost of the pipeline project is expected to be $1.2 billion, a rate of $1.67 million per mile or $1.99/mile per 84
U.S. EPA database of municipal landfills, http://www.epa.gov/epaoswer/nonhw/muncpl/landfill/index.htm#list 85 Myers, D.B., et al. “Cost and Performance Comparison of Stationary Hydrogen Fueling Appliances”, Task 2 Report, April 2002, prepared by Directed Technologies Inc. for The Hydrogen Program Office, Office of Power Technologies, U.S. Department of Energy. 86 DiPietro, J.P. and J.S. Badin. “Technology Characterizations for the E3 (Energy, Economics, Environment) Pathway Analysis.” 1995 U.S. DOE Hydrogen Program Review, DE95009296. 87 See Ref. 12. 88 http://www.midamerican.com/html/press_release.asp?RELEASE_ID=136
33
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
million Btu/day of capacity. The new pipeline is actually an expansion beside an existing pipeline, so the cost for right-of-way is not as high as it could be for a completely new pipeline. The capital cost for a hydrogen pipeline was assumed to be 40% higher on a lower heating value basis than for a natural gas pipeline, for a cost factor of $2.79/mile per million Btu/day of capacity. The cost for local hydrogen distribution pipelines was calculated in a different way and is addressed in Section 2.4.4. 2.3.8 Interstate Electricity Transmission The capital cost for electricity transmission lines was calculated from a curve fit to data for existing above-ground transmission lines over the range 60-230 kV and 32-1,060 MW.89 The cost per mile of transmission line is reasonably well represented by a linear function of capacity, as shown in Figure 2-7, with only a minor dependence on voltage. The cost for a 1,000 MW transmission line is calculated to be a base cost of $102,000 per mile plus a marginal cost of approximately $500 per MW-mile. 900 800
Cost ($1000/mile)
700 600 500 400 300 200 100 0 0
200
400
600
800
1000
1200
Capacity (MW)
Figure 2-7. Cost of high-voltage electricity transmission lines as a function of capacity.
2.3.9 Electrolyzers The cost for electrolyzers was assumed to be the long-term DOE goal of $300/kW. All electrolysis was assumed to take place at local hydrogen dispensing stations, corresponding to a rated electrolyzer peak capacity of 1,700 kWth (920 kg/day) with 89
U.S. DOE, Energy Information Administration, www.eia.gov/cneat/pubs_html/feat_trans_capacity/table2.html
34
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
average hydrogen production of 1,173 kWth (630 kg/day). Large-scale electrolysis near the electricity source with hydrogen transport in pipelines (analogous to centralized biomass gasification) was considered as a pathway but was calculated to be economically unfavorable compared to distributed electrolysis (see Section 3.4.2). 2.3.10 H2 Filling Station (Compression, Storage, and Dispensing) All of the potential pathways to hydrogen from renewable sources require a method of dispensing the hydrogen to vehicles. The dispensing station that we included has a compressor rated to 10,000 psig; storage at 10,000 psig for 12 hours of average hydrogen sales; and dispensing equipment to vehicles with 5,000 psig tanks. The capital cost for 920 kg/day (peak) compression, storage, and dispensing module was taken from the Directed Technologies estimate for distributed natural gas reforming.90 2.4 Hydrogen and Electricity Cost Models The costs for renewable energy were estimated using discounted cash flow (DCF) analysis. The DCF analysis incorporates the cost of capital investment, raw materials, and operating and maintenance costs to calculate a levelized cost of hydrogen or electricity. The financial and operational assumptions are different for each step in the process of delivering hydrogen to automobiles, so the specific values for important variables are addressed in the subsections below. The detailed DCF calculations for all scenarios are included in Appendix C. The assumptions common to all the cost calculations are an inflation rate of 3% applied to fixed operating costs, marginal corporate income tax rate of 38% (federal and state combined), and corporate overhead of 15% of revenues.
No special tax preferences were assumed for renewable hydrogen generation in the years 2030-2050. Any special tax credits that exist now are expected to expire by the time renewables mature into a multi-quad per year market. 2.4.1 Renewable Electricity Generation The assumptions about plant lifetime and capital payback period are especially significant factors in the cost of hydrogen since the capital cost contributes the bulk of the cost of production. For example, wind is “free” once the wind turbine is constructed. We have assumed that in 2030-2050 electricity generation in small units and that the generation equipment is owned by independent or investor-owned power producers. The plant lifetime and payback period were both assumed to be 25 years. The DCF parameters used for renewable electricity generation are listed in Figure 2-8, and the DCF parameters used for electricity generation from biomass gasification are listed in Figure 2-9.
Electricity generation from landfill gas was handled differently since the generation method is assumed to be natural gas turbines at the landfill site with the electricity distributed through the grid. The DCF parameters used for electricity generation from biomass gasification are listed in Figure 2-10. 90
See Ref. 85.
35
Hydrogen from Renewable Energy Sources
Cost of capital Capital recovery period
Directed Technologies Inc.
Value
Reference
10.8%
CECA modeling91 Power producers are unregulated utilities or IPP Fraction of time the electricity generation is operational
25 years
Availability factor
98%
Capacity factor
varies
Land lease rate
2.5% of sales
Variable operating costs Fixed operating costs Operator labor Working capital
See Section 2.1 U.S. Bureau of Land Management draft memorandum
0.5¢/kWh 2% of initial investment per year 3 full-time (each 12 hrs/day @ $40/hour)
Estimate based on multi-megawatt plants, not counting maintenance or corporate/overhead staff
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-8. Discounted cash flow parameters for renewable electricity generation (solar, wind, geothermal). Value Cost of capital Capital recovery period
10.8%
Reference Ref. 91
25 years
Availability factor
95%
Capacity factor
85%
Fraction of time the electricity generation is operational Estimate
Biomass conversion
80%
Typical value; residual is inorganic ash
Plant thermal efficiency
45%
Feedstock costs
varies
Operation and maintenance Fixed operating costs Operator labor Working capital
See Section 2.1.6
3% of initial investment per year 1.5% of initial investment per year 12 full-time (each 12 hrs/day @ $40/hour)
Insurance 1%, property taxes 0.5% Estimate
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-9. Discounted cash flow parameters for biomass gasifier electricity generation.
91
U.S. Dept. of Energy. The Comprehensive Electricity Competition Act: A Comparison of Model Results. Report#:SR/OIAF/99-04. CECA is an acronym for the Comprehensive Electricity Competition Act.
36
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Value Cost of capital
10.8%
Capital recovery period
20 years
Capacity
5.56 MW
Reference Ref. 91
Availability factor
95%
Fraction of time the electricity generation is operational
Capacity factor
85%
Estimate
Biomass conversion
80%
Typical value; residual is inorganic ash
Plant thermal efficiency
50%
LHV basis
Feedstock costs Operation and maintenance Fixed operating costs Operator labor
Working capital
$1.644/Mscf natural gas
See Section 2.1.6.4
3% of initial investment per year 1.5% of initial investment per year 1 full-time (each 12 hrs/day @ $40/hour) 30 days sales accounts receivable 15 days expenses accounts payable
Insurance 1%, property taxes 0.5% Estimate
Figure 2-10. Discounted cash flow parameters for landfill gas electricity generation.
37
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.2 Electrolysis The electrolysis cost includes capital recovery and operating and maintenance, but the cost of electricity to operate the electrolyzer is accounted for separately (see Section 2.4.1 above). The DCF parameters used for electrolytic production of hydrogen are listed in Figure 2-11. No credit was taken for sales of the oxygen that is co-generated with the hydrogen, although it would be possible to improve the economics of the electrolysis process by selling the oxygen. Value Cost of capital Capital recovery period Capacity
10.8%
Ref. 91
10 years 920 kg/day
Availability factor
100%
Capacity factor
69%
Electrolyzer efficiency
76%
Water
Reference
$2/1000 gallons
Fraction of time the electrolyzer is operational Accounts for daily and seasonal fluctuations in demand LHV basis Typical value for desalination
Operation and maintenance
1.0% of initial investment per year
Fixed operating costs
1.5% of initial investment per year
Insurance 1%, property taxes 0.5%
Operator labor
1 full-time (each 12 hrs/day @ $20/hour)
Labor for dispensing station is assumed to be a lower-skill task than plant operation
Working capital
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-11. Discounted cash flow parameters for electrolyzer distributed hydrogen generation.
38
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.3 Biomass Gasification and Reforming to Hydrogen Biomass gasification and reforming to produce pure hydrogen is most economical at the industrial scale. The gasification plants are considered to be long-life plants with the same characteristics as an electricity plant. The DCF parameters used for biomass gasification and reforming to hydrogen are listed in Figure 2-12. Value Cost of capital Capital recovery period Capacity
10.8%
Reference Ref. 91
25 years 110 metric tons H2/day
Availability factor
95%
Capacity factor
85%
Fraction of time the electrolyzer is operational Accounts for daily and seasonal fluctuations in demand
Plant thermal efficiency
45.2%
LHV basis
Biomass feedstock
varies
See Section 2.1.6
Operation and maintenance
3% of initial investment per year
Fixed operating costs
1% of initial investment per year
Electricity usage Operator labor
Working capital
Insurance 0.5%, property taxes 0.5%
1 kW/(kg H2/hr) 12 full-time (each 12 hrs/day @ $40/hour)
Estimate
10 days biomass inventory 30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-12. Discounted cash flow parameters for biomass gasification and reforming to hydrogen in large-scale centralized plants.
39
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.4 Electricity Transmission Lines and Hydrogen Pipelines Electricity transmission lines and interstate hydrogen pipelines were assumed to be longterm assets that have a lower cost of capital (i.e., lower risk) and longer payback period than process equipment. The DCF parameters used for the interstate electricity transmission and hydrogen pipelines are listed in Figure 2-13. Value Cost of capital Capital recovery period Capacity Availability factor Capacity factor Operation and maintenance Fixed operating costs Compressor fuel (pipeline only)
Operator labor
Working capital
Reference
8% 25 years 1,000 MW (electricity); 332.3 metric tons H2/hour (pipeline) 100% Electricity 70%; Pipeline 60%
Fraction of time the system is operational Accounts for daily and seasonal fluctuations in demand
1% of initial investment per year 1.5% of initial investment per year
Insurance 1.0%, property taxes 0.5%
0.15% of throughput Electricity 1 full-time per 300 miles, Pipeline 2 full-time per 100 miles (each 12 hrs/day @ $40/hour)
Estimate
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-13. Discounted cash flow parameters for interstate electricity transmission and hydrogen pipelines.
The cost of local hydrogen distribution by pipeline is assumed to be the same as the difference between the city gate and commercial prices of natural gas. The markup for distribution beyond the city gate includes capital recovery and continuing operating costs for the natural gas distribution company. In 2000, the U.S. natural gas average price at the city gate was $4.62/Mscf and to commercial customers $6.59/Mscf.92 The difference of $1.97/Mscf ($0.0102/kWh) was used as the local pipeline distribution cost for hydrogen.
92
U.S. DOE, Energy Information Administration. Natural Gas Annual 2000.
40
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2.4.5 Hydrogen Filling Stations (Compression, Storage, and Dispensing) All pathways for producing hydrogen from renewables include the final step of dispensing to vehicles. The cost for compression, storage, and dispensing modules were previously estimated by Directed Technologies Inc. in their analysis of distributed natural gas steam reforming.93 The DCF parameters used for the hydrogen filling station components are listed in Figure 2-14. Value Cost of capital Capital recovery period Capacity
Reference
10% 10 years 630 kg/day
Availability factor
100%
Fraction of time the station is operational
Capacity factor
69%
Accounts for daily and seasonal fluctuations in demand
Operation and maintenance
2% of initial investment per year
Fixed operating costs
1.5% of initial investment per year
Insurance 1.0%, property taxes 0.5%
Operator labor
1 full-time (each 12 hrs/day @ $40/hour)
Estimate
Corporate overhead
Working capital
0% of sales
Not included in filling station costs
30 days sales accounts receivable 15 days expenses accounts payable
Figure 2-14. Discounted cash flow parameters for hydrogen filling station compression, storage, and dispensing components.
2.4.6 Cost Summary The costs for each step in the hydrogen production chain are listed in Figure 2-15. The total cost of hydrogen for each pathway is calculated in Section 3.
93
See Ref. 85.
41
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
Cost Electricity Generation (¢/kWh electric) Wind class 4 Wind class 5 Wind class 6 PV solar Geothermal Dedicated energy crops Ag residual & wood waste Livestock manure MSW Landfill gas
4.72 4.41 3.84 9.80 3.87 6.89 6.21 5.15 5.95 3.11
Electricity Transmission (¢/kWh electric/mile)
1.78 x 10-3
Electrolysis (¢/kWh H2 LHV)
3.90
Hydrogen Production (¢/kWh H2 LHV) Dedicated energy crops Ag & wood waste Livestock manure MSW Long-distance H2 pipeline (¢/kWh H2 LHV/mile)
5.25 5.05 3.95 4.34 7.2 x 10-4
Local H2 pipeline (¢/kWh H2 LHV)
1.02
Local compression, storage, and dispensing (¢/kWh H2 LHV)
1.89
Landfill gas-distributed steam reforming Landfill gas cleanup (¢/kWh natural gas LHV)
0.606
Local natural gas pipeline (¢/kWh natural gas LHV)
1.10
Long-distance natural gas pipeline (¢/kWh natural gas LHV/mile)
5.1 x 10-4
Distributed natural gas reforming (¢/kWh natural gas LHV)
4.10
Figure 2-15. Cost factors used to calculate the cost of hydrogen in Section 3.
42
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
3 Proposed Hydrogen Generation Network 3.1 Development of Hydrogen Distribution and Cost Model A simulation was created to generate a hydrogen production scenario for the continental U.S. based on resource availability and cost. The estimated costs of hydrogen are based on generation and transportation costs, with interstate distances taken into consideration. The model strives to simulate an unregulated hydrogen market in which consumers in each state buy hydrogen generated from the least expensive available resource from any other state. The algorithm takes as input the resource availability, base costs, and needs in each state and outputs the resulting distribution of hydrogen resource usage and prices.
In the model, hydrogen is purchased in small units (0.0001 quads), with each state purchasing the cheapest available hydrogen in each buying round until all of that state’s hydrogen needs are met or the resources are consumed. At the beginning of each round, a cost for hydrogen generated using each resource from each state to each state is calculated. If a particular resource in a state is non-existent or has been consumed, that resource is unavailable for purchase. The price of each resource is the sum of generation and transmission costs, with conversion efficiencies taken into consideration. For electricity-producing resources, electricity is transmitted from the generation location to the grid, and consumed at the individual refueling stations where hydrogen is generated via electrolysis. For biomass resources that are reformed, hydrogen is piped from the production center through a longdistance pipeline (when used out-of-state) and distributed via a local hydrogen pipeline to refueling stations. The cost of long-distance transmission of hydrogen and electricity is calculated based on center-to-center interstate distances. For in-state consumption of resources, only the cost of local transmission is included in the price. For biomass resources, generation costs include the costs of the feedstocks, their collection, and costs associated with gasifying and reforming to hydrogen. The cost of the hydrogen is calculated on a lower heating value (LHV) basis, and later converted to price per mass. Excluding the reformation of landfill gas, each resource is used to generate either electricity or hydrogen at a large-scale facility in the state in which the resource is located. For methods employing electrolysis, the following equation provides the cost of hydrogen (in $/kWh): 1 CH 2 = (C G + CT D) + C E + C CSD η e (1 − lT ) where η e is the electrolyzer efficiency (0.76) lT is transmission line loss (0.06) C G is the cost of electricity generation (varies) CT is the cost of electricity transmission per mile D is the interstate distance for transmission (varies)
43
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
C E is the electrolyzer cost C CSD is the cost of hydrogen compression, storage, and dispensing
For methods employing reformation of biomass (excluding landfill gas), the following equation is used to calculate the cost of hydrogen:
CH 2 =
1 C G + C P − L + C P − D D + C CSD (1 − l P ) D
where l P is pipeline loss of H2 per mile (zero) C G is the cost of hydrogen generation (varies) C P − L is the cost of transmitting hydrogen through a local pipeline C P − D is the cost per mile of transmitting hydrogen through a long-distance pipeline D is the interstate distance for transmission (varies) C CSD is the cost of hydrogen compression, storage, and dispensing We assume that landfill gas, unlike the other reformation methods, is purified to methane and pumped into the existing natural gas pipeline. Steam reformation units located at the refueling stations then remove an equivalent amount of natural gas from the pipeline to generate hydrogen. The cost of hydrogen from reformed landfill gas is then given by the following equation:
CH 2 =
1 (C C + C P − L + C P − D D) + C R + C CSD η R (1 − l P ) D
where η R is the reformer efficiency (0.70) l P is pipeline loss of natural gas per mile (zero) C C is the cost of landfill gas cleanup C P − L is the cost of transmitting natural gas through a local pipeline C P − D is the cost per mile of transmitting natural gas through a long-distance pipeline D is the interstate distance for transmission (varies) C R is the cost of reforming C CSD is the cost of hydrogen compression, storage, and dispensing The result of these cost models is that the cost of hydrogen depends on the method being used to generate it and the distance between the state with the resource and the state where the hydrogen is consumed. Figure 3-1 shows the cost of hydrogen produced from each method when consumed locally and the additional cost of transmitting the energy 500 miles (an arbitrary distance chosen to be representative for interstate transmission).
44
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
DE pyrolysis (A) ManureR
ManureE LFgasR
LFgasE
MSWE MSWR
WoodWasteR
AgriResR
3
DEcropsR
GeoT
4
AgriResE
DEcropsE
Solar WC6
5
WC5
6 WC4
cost of hydrogen, ($/kg)
7
WoodWasteE
zero mile transmission 500 mile transmission
8
DE pyrolysis (B)
Hydrogen costs do not include profit at any step94. Naturally, in the cases where multiple methods of hydrogen production exist for a single resource, the cheapest method will be selected. As a result, since gasification/reformation is the most cost effective method of hydrogen production from biomass, neither electrolysis nor pyrolysis from biomass resources were considered in the model. Figure 3-2 provides a breakdown of the cost for hydrogen from both the electrolysis and reformation methods used in the model.
2 1 0 method
Figure 3-1. Hydrogen costs from the various production methods. For biomass resources, E refers to electrolysis, R to reformation.
7
a
Electricity Electrolysis Comp.,Stor.,Disp. Transmission (per 500 miles)
3.5
3
H2 cost, $/kg
6
H2 cost, $/kg
Feedstock Gas. & Ref. Local H2 pipeline Local NG pipeline Comp.,Stor.,Disp. Pipeline (per 500 miles)
b
5 4
2.5
2
1.5
3
GeoT
Solar
WC6
WC5
WC4
LFgasR
0
ManureR
0
MSWR
0.5
WoodWasteR
1
AgriResR
1
DEcropsR
2
Figure 3-2. Cost breakdown for hydrogen from (a) renewable electricity and (b) reformed biomass.
94
With the possible exception of profits for transmission of electricity or hydrogen pipelines.
45
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
In each buying round during the simulation, each state has the opportunity to buy hydrogen from the method-state combination that provides the lowest cost. Because resources are limited, states will be forced to buy from a variety of methods and other states to meet their hydrogen needs. The result is a distribution of hydrogen costs that varies from state to state. States at longer distances from low-cost resources and states with larger hydrogen demand will pay more, on average, for hydrogen. If a resource in a particular state is depleted mid-round, the remaining states in that round are still able to purchase that resource. This is done in order to prevent a bias due to purchase order in the round and results in errors of no greater than a few hundredths of a quad for any single resource. As each state fulfills its H2 need, it stops buying resources. Rounds continue until all of the states have stopped buying. Due to the remoteness of Hawaii and Alaska, neither of these states were considered to be connected to the continental United States via pipeline or electrical grid. Together, Hawaii and Alaska represent roughly 0.6% percent of the national demand for hydrogen. As a result, these states are considered special cases requiring individualized, relatively small-scale hydrogen generation and transport. The model, therefore, focuses on the continental U.S., with Hawaii and Alaska neither consuming hydrogen nor providing resources for its production. 3.2 Results The simulation generates a round-by-round history of the hydrogen purchase choices for all states, providing potential resource preferences, locations, and transmission. The output of the simulation represents neither an optimized lowest-cost hydrogen scenario nor a plan for hydrogen production and distribution, but rather a representative hydrogen infrastructure that can provide insights into the variables affecting the cost of producing and distributing renewably-generated hydrogen. In order to evaluate the model results, we can look at the statistics for hydrogen production on a national basis as well as examining the results for individual states. 3.2.1 National Statistics Figure 3-3 and Figure 3-4 show the annual resource availability and consumption from the model. Consumption of each resource depends both on its availability and on the cost of hydrogen from that resource relative to all other resources. While biomass and geothermal resources represent a small hydrogen potential relative to solar and wind energy, these resources make significant contributions to the 10-quad national hydrogen demand due to their relatively low cost. Solar energy, however, makes no contribution since it is considerably more costly than the abundant and widespread Class 4 wind energy with which it must compete.
As stated earlier, the cost of hydrogen in a particular state depends, in part, on the distance between that state and the location of the resource. For two renewable resources with comparable costs, the addition of transmission costs (whether pipeline or electrical) may make a nearby but more expensive resource desirable over a cheaper but distant resource. This can be seen in Figure 3-3 where Class 5 and Class 6 wind are not completely consumed, yet large quantities of hydrogen are generated using the more
46
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
expensive Class 4 wind. This indicates that some high quality wind resources (Class 5 and 6) are stranded, being too remote from heavily populated states to be cost-efficient for hydrogen production. Biomass 2.7 Quads
a
Biomass 2.7 Quads
b
GeoT 0.43 Quads
Solar 5.9 Quads
WC4 5.3 Quads
WC4 18 Quads
GeoT 0.43 Quads
WC6 1.7 Quads
WC6 0.98 Quads
WC5 3.1 Quads WC5 0.48 Quads
Figure 3-3. Model resource availability and usage (in quads of hydrogen per year). (a) Resource availability, totaling 31.9 quads. (b) Resource usage, totaling 9.95 quads.
Resource
Abbrev.
Wind Class 4 Wind Class 5 Wind Class 6 Photovoltaic Solar Geothermal Dedicated Energy Agricultural Residue Wood Waste Municipal Solid Waste Livestock Manure Landfill Gas
WC4 WC5 WC6 Solar GeoT DEcrops AgriRes WoodWaste MSW Manure LFGas
Cost ($/kg) $4.54 $4.40 $4.13 $6.91 $4.14 $2.84 $2.77 $2.77 $2.53 $2.40 $2.70
Potential (quads/yr) 18.1 3.1 1.7 5.9 0.43 0.57 0.81 0.73 0.18 0.29 0.13
Usage (quads/yr) 5.3 0.48 0.98 0.0 0.43 0.57 0.81 0.74 0.18 0.29 0.13
Figure 3-4. Summary of hydrogen cost (w/ 500 mile transmission), availability, and usage for each resource predicted by the model.
The average cost of hydrogen in the U.S. is predicted to be $3.98/kg ($0.12/kWh). The large cost difference between hydrogen from renewable electricity and from biomass results in a multimodal hydrogen cost distribution, as shown in Figure 3-5, with a clear gap between these types of resources. The national minimum and maximum are $2.29/kg ($0.07/kWh) for manure and $5.12/kg ($0.15/kWh) for Class 4 Wind, respectively.
47
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
2 1.8
H2 purchased (Quads/yr)
1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
2
2.5
3
3.5 4 Cost of hydrogen ($/kg)
4.5
5
5.5
Figure 3-5. National hydrogen cost distribution.
All else being equal, states with lower hydrogen demand are able to meet their needs at a lower average cost of hydrogen than states with large demands. This is because the model has each state purchasing the same quantity of hydrogen in each round. States with large hydrogen demands are forced to purchase quantities of expensive resources after small states have finished purchasing. This effect is displayed graphically in Figure 3-6, which demonstrates that the national average cost may decrease in consecutive rounds if states purchasing hydrogen at costs higher than the national average meet their needs and stop purchasing. Figure 3-7 provides the resulting average cost of hydrogen to each state. If, instead, all states were able to purchase the same fraction of their total demand in a purchase round, the average cost of hydrogen would be more uniform across states.
48
Directed Technologies Inc.
5
50
4.5
45
4
40
3.5
35
3
30
2.5
25
2
20
1.5
15
1
10
0.5
5
number of states purchasing
average cost of hydrogen ($/kg)
Hydrogen from Renewable Energy Sources
0 purchase round
Figure 3-6. Evolution of national average hydrogen cost throughout purchase process. Costs reflect hydrogen purchased in a given round and not the effect of hydrogen from previous rounds. AL AR AZ CA CO CT DE FL GA IA ID IL IN KS KY LA
$3.92 $3.72 $3.49 $4.09 $3.51 $3.25 $2.60 $4.72 $4.49 $3.11 $2.73 $4.16 $4.01 $3.23 $3.71 $3.63
MA MD ME MI MN MO MS MT NC ND NE NH NJ NM NV NY
$4.17 $4.11 $2.73 $4.17 $3.60 $3.84 $3.48 $2.60 $4.50 $2.54 $2.62 $2.72 $4.45 $3.08 $3.03 $4.52
OH OK OR PA RI SC SD TN TX UT VA VT WA WI WV WY
$4.30 $3.36 $3.33 $4.39 $2.62 $3.98 $2.56 $4.11 $4.03 $3.13 $4.34 $2.57 $3.68 $3.70 $2.64 $2.59
Figure 3-7. Average cost per kilogram of hydrogen for each state.
Generally, the least expensive hydrogen initially available to all states is from in-state biomass resources, followed by the biomass in neighboring states. Nearly half of all hydrogen from biomass is consumed in-state, with the amount transported out of state (via pipeline) decreasing with increasing interstate distance (see Figure 3-8). The average inter-state pipeline distance is 259 miles. As states consume their surrounding biomass resources, they are forced to purchase electricity for hydrogen generation, which may be transmitted over longer distances (see Figure 3-9). Only 24% of renewable electricity is used in-state for hydrogen production, and no obvious trend exists for the quantity of hydrogen transmitted as a function of transmission distance. The mean distance for renewable electricity transmission is 540 miles. It is interesting to note that
49
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
no resource is transmitted (via electrical lines or pipelines) more than 1500 miles, indicating that coastal states do not reach past the middle of the country to meet their hydrogen demand.
H2 piped (Quads/yr)
1.5
1.0
0.5
0
0
200
400
600
800 1000 distance (miles)
1200
1400
1600
1800
Figure 3-8. Distribution of transmission distances for hydrogen by pipeline. 1.8
H2 from electrolysis (Quads/yr)
1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0
0
500 1000 distance of electrical transmission (miles)
1500
Figure 3-9. Distribution of electricity transmission distances for electrolyzed hydrogen.
50
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
3.2.2 Selected State Statistics Because all forty-eight states in the model have access to the same resources, the combination of source states and resources any one state utilizes depends on the purchase decisions of all other states. Figure 3-10 through Figure 3-15 summarize the hydrogen purchases of four states: Texas, North Dakota, California, and New Jersey. These three states were selected as examples because they have very different demand and resource qualities. Figure 3-10 is a purchase history of the four states, showing the order of preference from available resources during the purchase process. Figure 3-11 summarizes the same data, emphasizing the source states and resources.
Texas has both a large hydrogen demand and a large supply of resources. It is able to meet a fraction of its demand with its own biomass resources and those of nearby states. However, as Figure 3-12a shows, the cost of hydrogen in Texas is dominated by its own wind resources. Texas supplies wind to other states as well, with the result that all of Texas’ Class 5 and 6 wind and most of Class 4 are consumed. North Dakota, like Texas, has significant wind resources. Unlike Texas, however, North Dakota has a low hydrogen demand and is distant from heavily populated states. North Dakota meets all of its hydrogen needs using biomass, both from within its borders and from other Midwestern and Plain states. As a result, the cost of hydrogen in North Dakota is comparatively low. According to the model, all of North Dakota’s wind is stranded, being too expensive for its own needs and too distant from states with high enough demands to justify the cost of transmission (see Figure 3-13b). For example, as Figure 3-10 and Figure 3-11 show, it is less costly for New Jersey to purchase Class 4 wind from as far as Minnesota than to purchase Class 5 wind from North Dakota (only one more state distant from New Jersey). California, like Texas, has both a large hydrogen supply and demand. In the purchase process, California continues to purchase hydrogen long after smaller states have met their hydrogen needs. This forces California to turn to resources which were too expensive for most other states, such as wind from Wyoming. As Figure 3-14 shows, the result is a lopsided distribution of hydrogen costs concentrated at more than $4.00/kg. The Eastern states have the disadvantage of having little local wind and, therefore, need to purchase wind electricity over long transmission distances for hydrogen production. New Jersey reaches as far as North Dakota for biomass resources before turning to the limited wind energy of Massachusetts, Maine, New York, New Hampshire, and Vermont (see Figure 3-10). This wind potential is quickly consumed, forcing New Jersey to purchase wind from Iowa and, finally, from Minnesota, which supplies more than half of New Jersey’s needs, making hydrogen relatively expensive (Figure 3-15a).
51
Hydrogen from Renewable Energy Sources
Source TX TX TX TX TX TX OK KS MO IA MN ND TX TX NM TX
TEXAS Resource quads/yr Manure 0.0096 MSW 0.0021 LFgas 0.0085 AgriRes 0.013 WoodWaste 0.0102 DEcrops 0.0126 DEcrops 0.0048 DEcrops 0.0016 DEcrops 0.0002 DEcrops 0.0007 DEcrops 0.0005 DEcrops 0.0002 WC6 0.0033 WC5 0.0331 WC6 0.0079 WC4 0.8655
$/kg $2.29 $2.42 $2.59 $2.65 $2.65 $2.72 $2.79 $2.83 $2.86 $2.91 $2.97 $2.98 $3.72 $3.99 $4.11 $4.13
Directed Technologies Inc.
Source CA CA CA CA CA OR WA NE NE NE SD KS ND CA CA NV CA CA UT NV NV WY
CALIFORNIA Resource quads/yr Manure 0.0059 MSW 0.0059 LFgas 0.0075 AgriRes 0.0036 WoodWaste 0.013 WoodWaste 0.0057 WoodWaste 0.0043 AgriRes 0.0024 WoodWaste 0.0001 DEcrops 0.0032 DEcrops 0.0088 DEcrops 0.002 DEcrops 0.0016 WC6 0.0245 GeoT 0.129 GeoT 0.192 WC5 0.0426 WC4 0.0619 GeoT 0.0419 WC5 0.0616 WC4 0.0633 WC6 0.5402
$/kg $2.29 $2.42 $2.59 $2.65 $2.65 $2.77 $2.83 $2.93 $2.93 $3.00 $3.01 $3.01 $3.02 $3.72 $3.73 $3.94 $3.99 $4.13 $4.15 $4.19 $4.34 $4.37
Source ND SD MN IA ND KS SD MN IA TX KS ND SD ND
NORTH DAKOTA Resource quads/yr Manure 0.0043 Manure 0.0013 Manure 0.0017 Manure 0.0001 MSW 0.0005 Manure 0.0002 MSW 0.0002 MSW 0.0009 MSW 0.0003 Manure 0.0001 MSW 0.0003 LFgas 0.0003 LFgas 0.0002 AgriRes 0.0152
$/kg $2.29 $2.33 $2.36 $2.41 $2.42 $2.44 $2.46 $2.49 $2.54 $2.55 $2.57 $2.59 $2.63 $2.65
Source NJ DE MD CT PA VA NC NJ MD PA VA NC OH MI IN TN NJ MD PA NJ NJ DE MD MD PA PA NJ VA NC MD SC VA KY WV IN IN GA OH IL IL KY IN TN IA IA IL MN WI MO IA MN ND MA ME MA NY NH VT PA MA NY NH ME ME IA MN
NEW JERSEY Resource quads/yr Manure 0.0003 Manure 0.0001 Manure 0.0004 Manure 0.0012 Manure 0.0006 Manure 0.0001 Manure 0.0007 MSW 0.0022 MSW 0.0004 MSW 0.0005 MSW 0.0003 MSW 0.0005 MSW 0.0003 MSW 0.0005 MSW 0.0002 MSW 0.0001 LFgas 0.0018 LFgas 0.0002 LFgas 0.0008 AgriRes 0.0002 WoodWaste 0.0009 AgriRes 0.0005 AgriRes 0.0013 WoodWaste 0.0011 AgriRes 0.0005 WoodWaste 0.003 DEcrops 0.0003 WoodWaste 0.0029 WoodWaste 0.006 DEcrops 0.0003 WoodWaste 0.0034 DEcrops 0.0001 WoodWaste 0.0011 DEcrops 0.001 AgriRes 0.0017 WoodWaste 0.0006 WoodWaste 0.0045 DEcrops 0.0003 AgriRes 0.0089 WoodWaste 0.0003 DEcrops 0.0003 DEcrops 0.0017 DEcrops 0.002 AgriRes 0.0045 WoodWaste 0.0002 DEcrops 0.0006 WoodWaste 0.0005 DEcrops 0.0021 DEcrops 0.0006 DEcrops 0.0007 DEcrops 0.0005 DEcrops 0.0002 WC6 0.0016 WC6 0.0007 WC5 0.0015 WC5 0.002 WC5 0.0005 WC5 0.0015 WC4 0.0061 WC4 0.0005 WC4 0.0036 WC4 0.0013 WC5 0.0007 WC4 0.0018 WC4 0.0414 WC4 0.146
$/kg $2.29 $2.31 $2.32 $2.32 $2.33 $2.35 $2.38 $2.42 $2.45 $2.46 $2.48 $2.51 $2.52 $2.56 $2.57 $2.58 $2.59 $2.62 $2.63 $2.65 $2.65 $2.67 $2.69 $2.69 $2.69 $2.69 $2.72 $2.72 $2.75 $2.75 $2.79 $2.79 $2.80 $2.80 $2.80 $2.80 $2.82 $2.82 $2.84 $2.84 $2.86 $2.87 $2.89 $2.90 $2.90 $2.91 $2.92 $2.93 $2.96 $2.97 $2.99 $3.06 $3.91 $4.12 $4.17 $4.23 $4.24 $4.24 $4.27 $4.32 $4.38 $4.38 $4.39 $4.53 $4.98 $5.06
Figure 3-10. Hydrogen purchases (in order of preference) for Texas, North Dakota, California, and New Jersey.
52
Hydrogen from Renewable Energy Sources
Source IA KS MN MO ND NM OK TX
TEXAS quads/yr Resource 0.0007 WC4 0.0016 WC5 0.0005 WC6 0.0002 DEcrops 0.0002 AgriRes 0.0079 WoodWaste 0.0048 Manure 0.9579 MSW LFgas
quads/yr 0.8655 0.0331 0.0112 0.0206 0.013 0.0102 0.0096 0.0021 0.0085
Directed Technologies Inc.
Source CA KS ND NE NV OR SD UT WA WY
CALIFORNIA quads/yr Resource 0.2939 WC4 0.002 WC5 0.0016 WC6 0.0057 GeoT 0.3169 DEcrops 0.0057 AgriRes 0.0088 WoodWaste 0.0419 Manure 0.0043 MSW 0.5402 LFgas
Source IA KS MN ND SD TX
NORTH DAKOTA quads/yr Resource 0.0004 AgriRes 0.0005 Manure 0.0026 MSW 0.0203 LFgas 0.0017 0.0001
quads/yr 0.1252 0.1042 0.5647 0.3629 0.0156 0.006 0.0231 0.0059 0.0059 0.0075
Source CT DE GA IA IL IN KY MA MD ME MI MN MO NC ND NH NJ NY OH PA SC TN VA VT WI WV
quads/yr 0.0152 0.0077 0.0022 0.0005
NEW JERSEY quads/yr Resource 0.0012 WC4 0.0006 WC5 0.0045 WC6 0.0468 DEcrops 0.0098 AgriRes 0.0042 WoodWaste 0.0014 Manure 0.0036 MSW 0.0037 LFgas 0.0032 0.0005 0.147 0.0006 0.0072 0.0002 0.0018 0.0057 0.0056 0.0006 0.0115 0.0034 0.0021 0.0034 0.0015 0.0021 0.001
quads/yr 0.2007 0.0062 0.0023 0.0107 0.0176 0.0245 0.0034 0.005 0.0028
Figure 3-11. Summary of source states and resources for hydrogen consumed by Texas, North Dakota, California, and New Jersey. 1.4
1.0 0.9
a
b
1.2
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
0.8 0.7 0.6
1
0.8
0.5
0.6
0.4 0.3
0.4
0.2
0.2 0.1
Manure
LFgas
MSW
4.2
WoodWaste
4
AgriRes
3.8
DEcrops
3.6
GeoT
3 3.2 3.4 Cost of hydrogen ($/kg)
Solar
2.8
WC6
2.6
WC5
2.4
WC4
0 2.2
Figure 3-12. Texas (a) hydrogen cost distribution, (b) resource usage. White bar shows unused hydrogen resource.
53
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
0.016
4.5
b
3.5
0.012
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
4
a
0.014
0.010
0.008
0.006
3
2.5
2
1.5
0.004
1
0.002
0
Manure
LFgas
MSW
WoodWaste
2.75
AgriRes
2.7
DEcrops
2.65
GeoT
2.45 2.5 2.55 2.6 Cost of hydrogen ($/kg)
Solar
2.4
WC6
2.35
WC5
2.3
WC4
0 2.25
0.5
Figure 3-13. North Dakota (a) hydrogen cost distribution, (b) resource usage. White bar shows unused hydrogen resource.
0.7
1
a
0.6
b
0.9
0.8
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
0.5
0.4
0.3
0.7
0.6
0.5
0.4
0.3
0.2
0.2 0.1
0.1 0
0
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
4.5
Solar
4
WC6
3 3.5 hydrogen cost ($/kg)
WC5
2.5
WC4
2
Figure 3-14. California (a) hydrogen cost distribution, (b) resource usage. White bar shows unused hydrogen resource.
54
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
0.20 5
0.18
a
x 10
-3
b
4.5
0.16 4
H2 supplied (Quads/yr)
H2 purchased (Quads/yr)
0.14
0.12
0.10
0.08
0.06
3.5
3
2.5
2
1.5
0.04 1
0.02 0.5
0
2
2.5
3
3.5
4
4.5
5
5.5
0
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Cost of hydrogen ($/kg)
Figure 3-15. New Jersey (a) hydrogen cost distribution, (b) resource usage. All resources were fully consumed.
3.3
Sensitivity Study for Important Variables
3.3.1 Hydrogen demand As we have seen, hydrogen produced by the gasification and reformation of biomass is significantly cheaper than hydrogen from electrolysis using renewable electricity. However, biomass resources are in short supply compared to the abundance of wind energy, particularly Class 4. The result is that hydrogen costs not only increase with increasing demand, but take a discontinuous jump once biomass resources are consumed. Figure 3-16 shows the national costs for hydrogen predicted by the model for national demands of 1, 3, 5, 10, and 15 quads/year while keeping costs and resource availability the same. As the demand increases, the cost of hydrogen becomes dominated by Class 4 wind. We can infer that hydrogen costs would take another jump when demand exceeded 26 quads/year, with Class 4 wind resources depleted and solar-produced hydrogen introduced to the market.
55
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
5.5
average H2 cost ($/kg)
5
maximum
4.5
4
mean
3.5
3
2.5
minimum 2
0
5
10 national H2 demand (Quads/yr)
15
Figure 3-16. Cost of hydrogen from renewables as a function of national demand for 1, 3, 5, 10, and 15 quads/year.
3.3.2 Low-cost solar Because Class 4 wind is so abundant and widespread, solar electricity must be costcompetitive with Class 4 wind electricity to play any significant role in the vehicular hydrogen market. The model was run for two low-cost solar scenarios: (1) the cost of solar electricity equals that of Class 4 wind, and (2) the cost of solar electricity equals that of Class 6 wind.
Scenario (1) represents a roughly 60% decrease in photovoltaic solar capital costs from the baseline model, or a 90% decrease from current costs. The result is that 0.71 quads/year of solar hydrogen are consumed, displacing some wind from all classes and some geothermal, and reducing the national average cost of hydrogen by two cents per kilogram to $3.96/kg. Figure 3-17a shows the national annual resource usage for this scenario. Scenario (2) represents a roughly 68% decrease in photovoltaic solar capital costs from the baseline model, or a 92% decrease from current costs. At this price, 2.1 quads/year of solar hydrogen are consumed, displacing wind and geothermal resources (see Figure 3-17b). The national average cost of hydrogen is reduced to $3.92/kg. Figure 3-18 summarizes the average price and solar resource usage from these alternate scenarios along with the baseline results.
56
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
a
Biomass 2.7 Quads
b
Biomass 2.7 Quads
WC4 4.1 Quads
WC4 5.2 Quads
GeoT 0.13 Quads
GeoT 0.39 Quads
Solar 0.71 Quads
WC6 0.44 Quads WC5 0.41 Quads
WC5 0.36 Quads WC6 0.44 Quads
Solar 2.1 Quads
Figure 3-17. Resource usage pies for scenarios in which solar electricity costs (a) equal wind Class 4 and (b) equal wind Class 6.
Cost of solar H2 w/ 500-mile transmission ($/kg)
$6.29 (baseline)
$3.92 (=WC4)
$3.51 (=WC6)
National average cost of H2 ($/kg)
$3.98
$3.96
$3.92
Solar H2 consumption (quads/year)
0.0
0.71
2.1
Figure 3-18. Results from scenarios with reduced costs for solar hydrogen.
3.3.3 Low-cost electrolyzed hydrogen The baseline model results reflect the high cost of hydrogen from electricity-generating resources compared to gasification/reformation of biomass. If renewable hydrogen from electrolysis could be made economically competitive with that from biomass, we might expect not only a significant decrease in the cost of hydrogen, but also a vastly different distribution for resource usage. Two scenarios were modeled in which hydrogen from Class 6 wind was competitive with biomass. Such a cost reduction would require both wind electricity and electrolysis costs to decrease, being unachievable through reductions in either component alone. The two scenarios examined assumed that, excluding transmission, hydrogen generated from Class 6 wind was at the same price as hydrogen from (1) dedicated energy crops and (2) livestock manure.95 The necessary cost reduction (in $/kg) to achieve these total costs for Class 6 wind were applied to the other electrolysis methods, as well.
95
The cost of hydrogen from Class 6 wind was actually made to be slightly more than expensive than the competing biomass, so that when choosing between biomass and Class 6 wind from in-state resources, the biomass was preferred.
57
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
In scenario (1), the national average cost of hydrogen was reduced to $3.25/kg from $3.98/kg. However, despite expectations, the national total consumption of each resource was relatively unchanged. The roughly 0.1 Quad/year increase in consumption of Class 6 Wind resources was taken from Class 4 Wind, indicating that previously stranded Class 6 wind resources had become economical compared to Class 4 resources. In scenario (2), the effect of scenario (1) was amplified, resulting in a national average cost for hydrogen of $2.91/kg. Figure 3-19 shows the resource utilization for both scenarios along with the baseline scenario. baseline
WC6 = DEcrops
WC6 = manure
20
a
Hydrogen from resource (Quads/yr)
18
b
c
16 14 12 10 8 6 4 2 0
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Figure 3-19. Resource uses and potential for three electrolyzed hydrogen costs. White bar shows unused hydrogen resource.
3.3.4 Low-cost electricity transmission As we have seen, the model predicts that significant Classes 5 and 6 wind resources are stranded due to the expense of electricity transmission. To investigate this effect, the model was run with two alternate electricity transmission costs: (1) 50% baseline transmission costs, and (2) 10% baseline costs.
Scenario (1) results in a national average hydrogen cost of $3.82/kg, while scenario (2) results in $3.64/kg. Figure 3-20 shows the national resource usage under each scenario, and Figure 3-21 shows summed Class 5 and 6 resource usage for states with those resources. baseline
50% transmission costs
10% transmission costs
20
a
Hydrogen from resource (Quads/yr)
18
b
c
16 14 12 10 8 6 4 2
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
Manure
LFgas
MSW
WoodWaste
AgriRes
DEcrops
GeoT
Solar
WC6
WC5
WC4
WC4
0
Figure 3-20. National resource usage and potential for three electricity transmission costs. White bar shows unused hydrogen resource.
58
Hydrogen from Renewable Energy Sources
State MA ME NH VT NY VA WV NC TX NM CO UT MT SD ND WY CA AZ NV WA OR ID TOTAL
Potential Quads/yr 0.012 0.014 0.007 0.006 0.005 0.002 0.004 0.008 0.249 0.178 0.671 0.056 1.248 0.033 0.699 1.290 0.067 0.010 0.103 0.057 0.009 0.058 4.788
Directed Technologies Inc.
Resource Usage (Quads/yr) Baseline 50% reduced 90% reduced cost transmission transmission 0.013 0.013 0.013 0.015 0.015 0.015 0.007 0.008 0.008 0.007 0.007 0.007 0.006 0.006 0.006 0.002 0.002 0.002 0.004 0.004 0.005 0.009 0.008 0.008 0.250 0.250 0.253 0.114 0.178 0.179 0.000 0.393 0.672 0.042 0.056 0.057 0.042 0.271 1.248 0.033 0.034 0.034 0.000 0.699 0.700 0.656 1.290 1.291 0.067 0.067 0.067 0.010 0.010 0.010 0.103 0.103 0.103 0.057 0.057 0.057 0.009 0.009 0.009 0.017 0.059 0.059 1.463 3.538 4.803
Figure 3-21. Summed Class 5 and 6 wind resource usage for reduced electricity transmission costs. (Resource usages greater than potential are due to model rounding errors.)
The model predicts that reducing electricity transmission costs by 50% will “unstrand” all Class 6 wind and some Class 5. Reducing by 90% unstrands all Class 5 and 6 wind. Of the states with Class 5 and 6 wind resources, Montana has the second greatest potential, but is also the most stranded of all states. In no case does the reduced transmission cost cut into biomass usage. 3.4
Alternative Scenarios
3.4.1 Electrolysis at centralized facilities While not assessed in the model, it may be possible to perform electrolysis on a large scale at centralized facilities located near the resources. Hydrogen would then be transmitted via long-distance and local pipelines to the consumers at refueling stations. The advantages of such a scheme would be decreased capital and operating costs for the electrolyzer (due to increased capacity and capacity factor) as well as decreased electricity transmission distances and costs. However, the costs for local and longdistance pipelining would be the same as that for biomass gasification. Compression, storage, and dispensing costs would be roughly the same.
59
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
For a given reduction in electrolyser costs, further reductions in the cost of hydrogen are linear with the transmission distance since hydrogen pipelines are significantly cheaper (per unit of energy transmitted) than electric power lines. The mean distance of electricity transmission for renewable electricity in the model is 554 miles. At this distance and assuming negligible electricity transmission costs for centralized electrolysis, electrolyser costs (per unit of energy produced) would need to decrease by roughly 11% for centralized electrolysis to be competitive with the distributed electrolysis used in the model. Such reductions may be reasonable for large electrolysis facilities, making centralized electrolysis a likely player in an electrolyzed-hydrogen infrastructure. 3.4.2 Electrolysis at location of resource Another possible method for producing hydrogen from renewable electricity is to perform the electrolysis at the location of the resource (wind, solar, etc.) and to pipe the hydrogen gas to the refueling stations. The advantages of such a plan are that longdistance electricity transmission lines could be eliminated and that the national hydrogen infrastructure might be made independent of the electricity infrastructure.
In spite of the intuitive advantages, several costly disadvantages make the at-resource electrolysis strategy for most resources non-competitive compared to electricity transmission for electrolysis at the refueling station. First, the at-resource electrolyzer must be sized to meet the full capacity of the electricity generating system so that no renewable energy is wasted. Secondly, due to mismatched hydrogen production and demand rates, considerable hydrogen storage would be necessary near the production site.96 Lastly, while long-distance pipelines are cheaper than electricity transmission, local pipeline costs make piping the more expensive transmission method at most interstate distances. Figure 3-22, below, provides a cost comparison of at-resource versus at-station electrolysis using electricity generated by Class 6 wind. This analysis is generous in favor of at-resource electrolysis because the selection of Class 6 wind takes advantage of Class 6’s high capacity factor compared to other renewable electricity resources, (excluding geothermal). Lower capacity factors mean not only higher electricity costs, but also larger and more expensive electrolyzers for at-resource electrolysis. Furthermore, the analysis ignores the cost of additional hydrogen storage capacity required with at-resource electrolysis. Despite these advantages, at-resource electrolysis still comes out more expensive than at-station electrolysis, $4.77/kg vs. $4.13/kg. Like many hydrogen production methods, at-resource electrolysis may be desirable for niche applications where access to an electrical transmission infrastructure is limited. 96
For example, considering seasonal wind fluctuations in Amarillo, Texas, hydrogen storage capacity would need to accommodate at least 12% of the annual production. For perspective, 12% of annual hydrogen production for automotive use in 2040 (1.2 quads) would be 3 trillion scf of hydrogen, equal to the U.S. current underground storage capacity for natural gas. The minimum required storage capacity for an installation can be estimated by calculating the maximum value at any time of the integral of hydrogen production rate minus hydrogen consumption rate ( max( ∫ ( production − consumption) dt ) for all t ).
60
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
However, because Class 6 wind can be utilized more cost-effectively through at-station electrolysis, this method is more likely to make a significant contribution to meeting the national demand for hydrogen and is the only method considered further in this report. Of the resources examined in this report, only geothermal with its high capacity factor could possibly receive an economic advantage through at-source electrolysis. However, that advantage is estimated to be small even when extra storage costs are ignored, $3.85/kg vs. $4.14, or 7%. Cost component Electricity
Electrolyzer
Electricity Transmission Long-distance pipeline Local Pipeline Compression, Storage, Dispensing Total ($/kWh) Total ($/kg)
Electrolysis site At-resource At-station $0.051 $0.054
Comments
$0.060
$0.039
N/A
$0.012
Assuming 76% efficient electrolyzer. Atstation electricity penalized for 6% transmission line loss. At-resource electrolyzer cost for 49% capacity factor on electricity production and 0.6 engineering scaling factor. PAR = PAS*(0.49)-0.6 Cost provided for 500-mile transmission.
$0.004
N/A
Cost provided for 500-mile transmission.
$0.010 $0.019
N/A $0.019
$0.143 $4.77
$0.124 $4.13
While considerable extra storage would be necessary for at-resource H2 production, no penalty is applied here.
Figure 3-22. At-resource versus at-station electrolysis cost comparison (in $/kWh) for hydrogen production from Class 6 Wind.
3.4.3 Electricity from Renewables, Hydrogen from Natural Gas The goal of renewable electricity is presumably to decrease pollution and greenhouse gas emissions and to enhance energy security, rather than to provide the most economical source of electricity. All of the renewable resources investigated in this study can be used to generate electricity, although the costs of the production vary widely. Figure 3-23, below, shows that wind and geothermal resources produce electricity that is less costly than that from biomass. Here we will briefly investigate the relative economics and CO2 emissions from replacing some nuclear and fossil fueled electricity with wind and geothermal resources while producing hydrogen for transportation from natural gas.
61
Directed Technologies Inc.
Solar
Hydrogen from Renewable Energy Sources
ManureE
LFgasE
MSWE
WoodWasteE
GeoT
WC6
6
WC5
8 WC4
cost of electricity, (cents/kWh)
10
AgriResE
generated cost line loss 500 mile transmission DEcropsE
12
4
2
0 method Figure 3-23. Costs of renewably generated electricity.
Currently, 6-7% of the 12 quads of electricity consumed in the United States is generated from renewable sources. This percentage has held relatively steady for the past 50 years.97 The U.S. Department of Energy expects electricity consumption to rise by roughly 35% by 2025, but anticipates little change in the fraction supplied by renewables.98 If we take into consideration the 76% efficiency for electrolysis applied in this report, the 7 quads of hydrogen from geothermal and wind energy is equivalent to 9 quads of electricity. This quantity may be as much as one third of the annual electricity consumption by the year 2040. It is doubtful whether the United States electrical infrastructure could support this high fraction of renewable energy without grid stability problems. However, for the sake of this brief analysis, we will assume that 9 quads/yr of renewable electricity is possible. The estimated national annual cost and CO2 emissions for this scenario is compared to the model baseline in Figure 3-24. Based on the prices of electricity used in this report (projected for 2040) and an average wholesale cost for electricity of $0.03/kWh from non-renewable production methods,99 9 quads/yr of renewable electricity is estimated to 97
Annual Energy Review 2001: Energy Overview: 1949-2001. Energy Information Administration. http://www.eia.doe.gov/emeu/aer/pdf/pages/sec8_5.pdf 98 Annual Energy Outlook 2003 with projections to 2025. Energy Information Administration. http://www.eia.doe.gov/oiaf/aeo/pdf/appa.pdf 99 This cost is generously low to account for advances in conventional electricity generation.
62
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
cost the United States an additional $40billion annually and to alleviate roughly 1500Mtonnes/yr100 of CO2 emissions generated from the current mix of coal, natural gas, and nuclear electricity. The average cost of hydrogen from renewable electricity resources in the year 2040 was estimated to be $4.49/kg, or $280 billion per year for 7 quads. The estimated cost for the same hydrogen reformed from natural gas is estimated to be roughly $120 billion annually, with a CO2 release of 600 Mtonnes/year101. The result of this alternate scenario is roughly 900Mtonnes/year additional CO2 reduction for $235 billion/year less when compared to the model baseline. (For perspective, total U.S. CO2 emissions from all human sources in 2000 were estimated at 5800 Mtonnes.102) Scenario
BASELINE: Renewable H2; Fossil-based elec. 7 quads/yr renewable elec.; NG-based H2
Cost of 9 quads electricity compared to baseline ($/yr)
$0 (from coal, natural gas, nuclear) $40billion (from wind and geothermal)
Cost of 7 quads hydrogen ($/yr)
$395 billion (from wind & geothermal) $120 billion (from steam methane reformation)
CO2 emissions compared to baseline (Mtonnes/yr)
0 -1500 + 600 = -900
Figure 3-24. Estimated cost and CO2 emissions of alternative renewable electricity and hydrogen from natural gas compared to the model baseline.
In reality, each renewable energy source has its own merits and disadvantages for each application, considering storage, stability, and distribution. However, we generalize that if faced with a choice between using a quantity of renewable electricity as grid electricity or hydrogen, it is both less expensive and less greenhouse gas intensive to generate electricity with renewables and produce hydrogen from fossil fuels than vice-versa. 3.4.4 Nuclear Hydrogen While not addressed in the model, methods for producing hydrogen at nuclear plants have been proposed. Nuclear power has the advantages of zero greenhouse gas emissions, relatively low cost compared to renewable electricity (at least in the near-term), and excess heat. Because the electrical efficiency of electrolysis improves at elevated temperatures, nuclear’s excess heat provides a lower cost, higher efficiency method for electrolytic production of hydrogen.
Another option (besides electrolysis) for nuclear production of hydrogen is thermochemical processing, in which water is split into hydrogen and oxygen via a set of 100
Based on CO2 emissions of 245, 134, and zero Mtonnes/quad of electricity and a 50:35:15 mix for coal, natural gas, and nuclear electricity, respectively. 101 Based on a 70% efficient steam methane reforming process producing hydrogen for $2/kg. 102 Emissions of Greenhouse Gases in the United States 2000. Energy Information Administration. November 2001.
63
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
chemical reactions with heat supplying all of the energy. One such process developed by General Atomics produces H2SO4 and HI as intermediary species in the process, with sulfur and iodine recycled in the process. Thermal efficiencies of roughly 50% are achievable103, but little or no electricity is consumed in the process. 3.5 Study Limitations The results reported above are based on reasonable estimates of capital and operating costs taken from the literature. We did not seek to optimize each source of hydrogen nor did we seek to design new processes. Some of the technology and cost projections are speculative since predictions about costs forty years from now have a high degree of uncertainty.
We started with the assumption that 10 quads per year of hydrogen would be used for transportation purposes and that all of that hydrogen would be derived from renewable resources. Those assumptions likely resulted in a sub-optimal solution for the U.S. energy market as a whole, especially with respect to the cost of hydrogen. For example, natural gas may be used to produce low-cost hydrogen, freeing up renewable resources for residential, commercial, and industrial electricity (see Section 3.4.3). The study assumed that the 10 quads of vehicular hydrogen from renewable resources would not need to compete with hydrogen from non-renewable sources. Non-renewable hydrogen would likely be produced from either the steam methane reformation (SMR) of natural gas or from electrolysis using electricity produced primarily from coal, natural gas, and nuclear energy. Figure 3-25 compares hydrogen costs for renewable hydrogen with similar processes using non-renewable resources. Reformation of natural gas for hydrogen production may take place at large plants or at refueling station installations, and representative costs for each are included in the figure. We assume that electrolysis always takes place at the refueling station. A given process using a non-renewable energy source is generally less expensive than its renewable alternative, but reformation is considerably less expensive than electrolysis (~25-50%), regardless of the energy sources used. Had renewable hydrogen been required to compete with some quantity and distribution of non-renewable sources, a different infrastructure with a lower average cost of hydrogen could be expected.
103
Schultz, Ken. Efficient Production of Hydrogen from Nuclear Energy (presentation). California Hydrogen Business Council. June 27, 2002.
64
Hydrogen from Renewable Energy Sources
H2 Production Method
Reformation processes
Electrolysis processes
Directed Technologies Inc.
Cost Range ($/kg H2)
Comments
Large plant SMR
$2.21 $2.63
Based on existing 327 and 33 tonne/day facilities running on natural gas w/ 500mile hydrogen pipeline and natural gas cost of $4.17/106 Btu (average of industrial and utility prices in 2000)
Station-size SMR
$1.99 $3.38
Based on proposed 920 and 115 kg/day installations running on natural gas
Large plant biomass
$2.40 $2.84
Conventional electricity
$3.74 $4.67
Based on proposed 108 tonne/day facilities running on biomass resources w/ 500-mile hydrogen pipeline Based on electricity at wholesale costs of 3-6 cents/kWh w/ 500-mile transmission
Renewable electricity
$4.13 $4.54
Based on projected costs for geothermal and wind electricity w/ 500-mile transmission
Figure 3-25. Hydrogen costs from renewable and non-renewable sources, including transmission, compression, storage, and dispensing costs.
65
Hydrogen from Renewable Energy Sources
Directed Technologies Inc.
4 Conclusions The annual generation of 10 quads of hydrogen in the years 2030-2050 from renewable sources for transportation uses in the U.S. is technically achievable and, according to our model, leads to a national average hydrogen cost of $3.98/kg ($33.24/GJ, LHV basis), excluding profits. If hydrogen fuel cell vehicles achieve the projected efficiency (2.2times higher than conventional internal combustion vehicles), the cost of hydrogen would be equivalent to $1.81 per gallon for gasoline untaxed and without profit for the refiner or the distributor.104 The state average costs of renewable hydrogen vary widely ($2.54$4.72/kg), depending on the availability of resources and the necessary transmission distances. Wind and biomass are the most significant resources (on an energy supplied basis) for hydrogen production, with geothermal playing a small role due to its limited potential. A significant quantity of high-quality wind resources (Classes 5 and 6) in the central U.S. will be stranded due to the prohibitive cost of transmission. Hydrogen from renewable electricity is expensive compared to that from the reformation of biomass for three important reasons. First, most renewable electricity, especially wind and solar, is expensive compared to electricity from fossil fuels due primarily to high capital costs for the wind and solar installations with relatively low capacity factors (all
