Regularities in Early Hydrogen Station Size Distributions

Transcription

1 Regularities in Early Hydrogen Station Size Distributions Marc W. Melaina and Joel Bremson Hydrogen Pathways Program Institute of Transportation Studies, UC Davis 26th USAEE / IAEE North American Conference September 24-27, 2006 Ann Arbor, Michigan

2 Hydrogen Pathways Program 4-Year Research Program focusing on hydrogen Directors: Dan Sperling and Joan Ogden Interdisciplinary, with focus on engineering-based models New research program will begin in 2007 Sustainable Transportation Energy Pathways (STEPS) Expansion in Scope: 1) Hydrogen, 2) Biofuels, 3) Electric-drive Program Website: Melaina and Bremson, USAEE/IAEE

3 Presentation Overview Challenges for Early Hydrogen Station Networks Relative Station Size Distributions Size Distributions for Early Hydrogen Stations Cluster Analysis of Gasoline Stations Cost Implications Example Melaina and Bremson, USAEE/IAEE

4 Early Hydrogen Station Networks Chicken and Egg Problem Introductory phase will include a small number of underutilized stations serving early markets Investments may occur through partnerships Modeling Approach Evaluate hydrogen infrastructure costs and investments patterns on city and regional scales (bottom-up) Why are station sizes important? Economies of scale for hydrogen infrastructure components Station sizes represent the spatial distribution of demand Distribution of demand influences return on investment Goal of improving realism of models Melaina and Bremson, USAEE/IAEE

5 Early Networks will be Sparse and Competitive? Wide range of station sizes Growing markets subject to demand fluctuations Late entrants may have cost advantages How might infrastructure investments change over time? Birmingham with early station network Melaina and Bremson, USAEE/IAEE

6 Analysis of Gasoline Station Data in Multiple Cities MPSI survey data: 1) average monthly output, 2) Lat/Long Various city sizes, densities and regions Quantitative basis for representing service of refueling Stations per sq. mi > 3.0 M M M M < 0.25 M 0 1,000 2,000 3,000 4,000 5,000 6,000 Population Density (persons per sq. mi.) Population < 0.25 m m m m > 3.0 m Melaina and Bremson, USAEE/IAEE

7 Station Size Distributions Ranked left to right by station size (average monthly output in 1000s gal/month) 500 Average Output (1000s gal/mo) San Diego, CA 2000 Nashville, TN 2003 Minneapolis St. Paul, MN Number of Stations Melaina and Bremson, USAEE/IAEE

8 Relative Station Size Distributions Normalize by average station size (y axis) and total number of stations (x axis) 400% Relative Station Size 350% 300% 250% 200% 150% San Diego, CA 2000 Nashville, TN 2003 Minneapolis St. Paul, MN 1999 All Stations (ranked) 100% 50% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of all Stations Melaina and Bremson, USAEE/IAEE

9 Similar Distribution for all Cities 400% 400% Percent of Average Station Size 350% 300% 250% 200% 150% 100% > 3.0 million Los Angeles, CA 2000 Atlanta, GA S.F. Bay Area, CA 2002 All Stations Percent of Average Station Size 350% 300% 250% 200% 150% 100% million Phoenix, AZ San Diego, CA 2000 Minneapolis St. Paul, MN 1999 Denver, CO 2000 Kansas City, MO 2000 San Jose, CA 2002 All Stations 50% 50% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of Total Stations 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of All Stations Percent of Average Station Size 400% 350% 300% 250% 200% 150% 100% million Austin, TX Nashville, TN 2003 Salt Lake City, UT Concord-Livermore, CA Hartford, CT Birmingham, AL 1999 All Stations 50% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of Stations Melaina and Bremson, USAEE/IAEE

10 Similar Distribution for all Cities 400% 400% Percent of Average Station Size 350% 300% 250% 200% 150% 100% million Colorado Springs, CO 2000 Vallejo-Fairfield, CA 2002 Wichita, KS 1999 Bakersfield, CA, 2001 Santa Rosa, CA 2002 Chattanooga, TN 1999 Gulfport-Biloxi, MS 2003 All Stations Percent of Average Station Size 350% 300% 250% 200% 150% 100% million Marin, CA 2002 Gainesville, FL 2000 Santa Cruz, CA 2002 Salinas, CA Olympia, WA 1999 Fort Collins, CO 2000 Bloomington, IL 2000 All Stations 50% 50% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of Stations 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of Stations 400% 350% < 0.1 million Monterey, CA 2002 Hattiesburg, MS Percent of Average Station Size 300% 250% 200% 150% 100% Dalton, GA 2000 Cleveland, GA 2000 Owensboro, KY 2003 All Stations 50% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of Stations Melaina and Bremson, USAEE/IAEE

11 Station Cluster Analysis Clustering Algorithm 1. Compute linear distance between all station pairs 2. Find minimum distance pair 3. Eliminate smaller station and allocate it s average output volume to large station 4. Repeat steps 2 & 3 until one station remains Results in a maximum coverage network, preserving the (relative) demand at each station Melaina and Bremson, USAEE/IAEE

12 Example: Nashville, TN with 2 mile Cluster Distance 15 miles Melaina and Bremson, USAEE/IAEE

13 Network Size vs. Cluster Distance Number of stations drops off rapidly with increased cluster distance percent of stations eliminated at a cluster distance of 1 mile percent at 2 miles Percent of Stations Remaining Percent of Stations Remaining 100% Atlanta, GA 90% Austin, TX Salt Lake City, UT 80% San Jose, CA 70% Concord-Livermore, CA Minneapolis-St. Paul, MN 60% San Diego, CA Denver, CO 50% Hartford, CT 40% S.F. Bay Area, CA Phoenix, AZ 30% Kansas City, MO Nashville, TN 20% 10% 0% Cluster Distance (miles) 100% Bakersfield, CA 90% Santa Rosa, CA Olympia, WA 80% Marin, CA 70% Fort Collins, CO Gulfport-Biloxi, MS 60% Gainesville, FL Salinas, CA 50% Colorado Springs, CO Bloomington, IL 40% Vallejo-Fairfield, CA 30% Santa Cruz, CA Chattanooga, TN 20% Birmingham, AL 10% 0% Cluster Distance (miles) Melaina and Bremson, USAEE/IAEE

14 Size Distributions are Preserved with Clustering Same relative station size distribution is observed across a range of cluster distances Distributions for smaller cities deviate more from average distribution Percent of Average Station Output Percent of Average Station Size Percent of Total Stations Percent of Total Stations Atlanta Austin Concord-Liv. Denver Hartford KansasCity,MO Minn.St.Paul Nashville Phoenix SaltLakeCity SanDiego SanJose,CA SF.Bay.Area Fit to Original Stations Colorado Springs, CO Fort Collins, CO Gainesville, FL Gulfport-Biloxi, MS Marin, CA Olympia, WA Salinas, CA Santa Cruz, CA Santa Rosa Birmingham, AL Bloomington, IL Chattanooga, TN Bakersfield Vallejo-Fairfield, CA Fit to Original Stations Melaina and Bremson, USAEE/IAEE

15 Result is Simulation of Early Network Expansion Cluster Distance = 4.0 miles Melaina and Bremson, USAEE/IAEE

16 Result is Simulation of Early Network Expansion Cluster Distance = 3.0 miles Melaina and Bremson, USAEE/IAEE

17 Result is Simulation of Early Network Expansion Cluster Distance = 2.0 miles Melaina and Bremson, USAEE/IAEE

18 Result is Simulation of Early Network Expansion Cluster Distance = 1.6 miles Melaina and Bremson, USAEE/IAEE

19 Result is Simulation of Early Network Expansion Cluster Distance = 1.3 miles Melaina and Bremson, USAEE/IAEE

20 Result is Simulation of Early Network Expansion Cluster Distance = 1.0 miles Melaina and Bremson, USAEE/IAEE

21 Cost Example: Onsite SMR for all Stations Onsite steam methane reformer systems have strong economies of scale Used here as extreme example of effects of station size distributions In reality, various hydrogen production, storage and delivery methods would be employed Others: liquid or gaseous tank truck, electrolysis, pipeline Total Capital Cost ($ millions) Levelized Cost of Hydrogen $12 $10 $8 $6 $4 G. Storage-Based Onsite SMR Onsite electrolysis NAS SMR Current NAS SMR Future y = x R 2 = y = x R 2 = $2 y = x R 2 = 1 $ ,000 1,250 1,500 1,750 2,000 2,250 2,500 2,750 $10 $9 $8 $7 $6 $5 $4 $3 $2 $1 y = x R 2 = y = x R 2 = 1 Station Capacity (kg/day) y = x R 2 = $ ,000 1,500 2,000 2,500 3,000 Station Capacity Onsite SMR Onsite Electrolysis GH Storage Melaina and Bremson, USAEE/IAEE

22 Examine Cost in 3 Hypothetical Cities Two city sizes First has half the population density of the other two Range of station densities (stations per square mile) City A City B City C Population 1,000,000 Population 1,000,000 Population 2,000,000 Sq. miles 529 Sq. miles 265 Sq. miles 500 Pop. Density 1,889 Pop. Density 3,778 Pop. Density 3,996 Total stations 353 Total stations 235 Total stations 458 Stations/sq mi 0.67 Stations/sq mi 0.89 Stations/sq mi 0.92 % H2 market 33% % H2 market 33% % H2 market 17% Number of Stations Station Density City A % of Existing Stations Station Density City B % of Existing Stations Station Density Melaina and Bremson, USAEE/IAEE City C % of Existing Stations (Stns/sq mi) (N H2 / N gsln ) (Stns/sq mi) (N H2 / N gsln ) (Stns/sq mi) (N H2 / N gsln ) % % % % % % % % %

23 Hydrogen Station Sizes Can determine station sizes for the hydrogen station networks serving each hypothetical city Example below: 250,000 kg/day for entire network Station Capacity (kg/day) 12,000 10,000 8,000 6,000 4,000 2, Stns 50 Ave 100 Stns 100 Ave 150 Stns 150 Ave Number of Stations 2,000 1,800 1,600 1,400 1,200 1, Average Refills per Day Melaina and Bremson, USAEE/IAEE

24 Resulting Levelized Cost of Hydrogen Variation in levelized costs are due to economies of scale at the station level Levelized costs per station below are for networks providing 250,000 kg/d of hydrogen 5.5 Levelized Cost of Hydrogen ($/kg) Stns 50 Ave 100 Stns 100 Ave 150 Stns 150 Ave Number of Stations Melaina and Bremson, USAEE/IAEE

25 Challenge: Provide Adequate Refueling Availability while Minimizing Average Network Cost Increase in number of stations has a strong effect on average networks costs (especially if all stations use onsite SMR technology) Analysis suggests that various technologies may be employed to suit range of station sizes New methods needed to determine realistic investment behavior over time Average Network Cost of Hydrogen ($/kg) $5.50 $5.00 $4.50 $4.00 $3.50 $3.00 $ Number of Stations Total Network 5000 kg/d 10,000 kg/d 30,000 kg/d 50,000 kg/d 100,000 kg/d 250,000 kg/d 500,000 kg/d Melaina and Bremson, USAEE/IAEE

26 Questions, Comments, Suggestions? Contact Info Dr. Marc W. Melaina Melaina and Bremson, USAEE/IAEE

27 Levelized Costs by Station Size Cost Element Units Onsite SMR Onsite Electrolysis Mini Small Med. Large Mini Small Large Capital Costs Elements Direct capital kg/yr $2.97 $2.12 $1.38 $1.10 $2.50 $1.77 $1.09 Indirect capital kg/yr $1.29 $0.92 $0.73 $0.58 $1.09 $0.77 $0.58 Total Capital Costs (subtotal) kg/yr $4.26 $3.04 $2.11 $1.68 $3.60 $2.54 $1.67 Variable Costs Elements Natural Gas $/kg $1.27 $1.27 $1.09 $ Electricity $/kg $3.79 $3.79 $2.61 Labor $/kg $0.62 $0.42 $0.21 $0.12 $0.62 $0.42 $0.21 Other Variable Costs $/kg $1.01 $0.87 $0.76 $0.70 $0.91 $0.79 $0.63 Total Variable Costs (subtotal) $/kg $2.90 $2.56 $2.05 $1.73 $5.32 $5.00 $3.45 Levelized Cost Parameters Interest rate %/yr 10.0% 10.0% 9.0% 9.0% 10.0% 10.0% 9.0% Capital Recovery Factor %/yr 13.1% 13.1% 11.0% 11.0% 13.1% 13.1% 11.0% Years operation year Total Levelized Cost of H2 $/kg $7.16 $5.60 $4.16 $3.41 $8.92 $7.53 $ Melaina and Bremson, USAEE/IAEE