Monday, December 22, 2008

Purple Line Alternatives and Greenhouse Gas Emissions

This guest blog is by David Salzman. I realize many are worn out from the Purple Line debate but let's keep it civil in the comments. Of course, a variety of viewpoints are expected and welcome on this controversial issue.

This posting provides an apples-to-apples comparison of the medium-investment light-rail transit (LRT) and low-investment bus rapid transit (BRT) vehicles competing for consideration in the inner Purple Line alignment. It analyzes their greenhouse gas performance with all the backup numbers, sources, and calculations spelled out in human-readable and spreadsheet-readable versions.

My objective is to facilitate further analysis, particularly reasoned comments explaining any disagreements with my data or methodology.

Since this blog is ultimately about politics, comments regarding the alignment of public policy with this quantitative evidence (or persistence of policy assertions in contravention of the evidence) are welcomed too.

The Bottom Line:

All of the electrified light-rail transit (LRT) options would emit far more carbon dioxide than displaced from riders' personal cars. In fact a 60-seat light-rail car (TCar) would need to pull 75 passengers from their cars to reach CO2 break-even. The most vigorously pursued option, a medium-investment LRT, would generate 73,049,938 net pounds of carbon dioxide annually (113,423,818 emitted minus 40,373,880 displaced from its new riders' cars), or 25.6 net pounds added per mile traveled per rail car (TCar).

In contrast, all of the bus rapid-transit (BRT) options would pull enough riders out of their cars to yield less total production of carbon dioxide. The most vigorously pursued option, a low-investment BRT running on compressed natural gas (CNG), would save 14,380,310 net lbs of carbon dioxide annually (9,591,681 emitted minus 23,971,991 displaced from its new riders' cars), or 6.61 net lbs of carbon dioxide removed per mile traveled per rapid-transit bus. The BRT system only needs to pull 8 passengers from their cars to reach CO2 break-even. By the Draft EIS's numbers, each BRT vehicle pulls 13 passengers from their cars.

I have imputed a schedule for the LRT from the table of service shown on page 3-3 of the Draft EIS. (Note that the LRT only delivers or collects 1800 passengers per hour each way at Bethesda during the morning rush hour, despite twice that capacity being needed for the document's own figures to tie together.) The notation "6 2-TCar" means six trains per hour, each comprising two 60-seat electrified rail cars.

Weekdays:
5:00 - 7:00 AM 6 1-TCar trains per hour,
7:00 - 9:30 10 3-TCar,
9:30 - 4:00 PM 6 2-TCar,
4:00 - 7:00 10 3-TCar,
7:00 - 9:00 6 2-TCar,
9:00 - midnight 6 1-TCar

Saturday:
12:00 - 3:00 AM 3 1-TCar,
7:00 - 9:30 3 1-TCar,
9:30 - 7:00 PM 6 1-TCar,
7:00 - 9:00 2 2-TCar,
9:00 - midnight 3 1-TCar

Sunday:
12:00 - 3:00 AM 3 1-TCar,
7:00 - 9:30 3 1-TCar,
9:30 - 9:00 PM 6 1-TCar,
9:00 - midnight 3 1-TCar

This totals to 3372 TCar-hours per week, or 1.94 TCars per average train. 68.4% are off-peak.

The table grants the LRT an average passenger load of 89.1% in order to avoid the pollution from less-full trains, though this makes the estimated off-peak ridership absurdly high, e.g. more than 40 passengers all the way until 3:00 AM Saturday and Sunday mornings. The actual service level will consequently need to be higher, especially 5:00 - 7:00 AM weekday mornings, but that would disadvantage the LRT option further and I wanted to give it every reasonable benefit of the doubt since the hard numbers prove it to be such a gawdawful choice anyway.


NOTES for the spreadsheet below:

+ Two versions follow. The first is a straightforward table with all the data sources spelled out and entries explained; the second allows you to recreate the spreadsheet yourself and then waste countless hours trying to contrive a way to make the LRT less awful an environmental choice.

+ The carbon dioxide offset for getting people out of their cars onto rapid transit as new riders will diminish over time, though I don't include the effect. Passenger car technology improves faster than BRT technology, and BRT technology improves faster than LRT technology. In any case, cars turn over every 5 - 7 years on average, so the car fleet adopts gas-saving technology soonest.

+ Only "new trips" save gas by getting people out of their cars. The highest fraction of riders the MTA could stretch itself to claim would be new is 30.7% for the medium-investment LRT and 28.5% for the low-investment BRT. Everybody else would already be taking mass transit anyway, so they save no carbon offsets by switching to the new system. THIS IS AN IMPORTANT POINT.

+ I have not penalized the LRT for spending more carbon dioxide than the bus seat it pulls non-new riders out of, but probably should.

+ Be careful when interpreting the % occupancy figures. 100% occupancy means that every seat is occupied 3 1/2 times on average along the 32-mile circuit, since the average trip length is only 8.92 miles. The Draft EIS calculates the displacement of passenger car vehicle miles traveled (car VMT) for the high-investment LRT option (20,500 new riders), so I have derated the savings for the medium-investment LRT (19,200 new riders) and low-investment LRT (11,400 new riders). But the average trip length remains 8.92 miles whether the rider travels in a passenger car, BR bus, or LRT rail car.

+ I have set the expected MPG per average car in 2030 to the 2007 number of 22.4. Yes, it should be twice as high by then, and the CAFE standard for 2020 is 35 MPG, but I was trying to give the LRT every reasonable advantage to avoid the finding that it is browner and dirtier than cars, where the BRT is greener and cleaner.

+ There should be slight corrections for the vehicles getting heavier with more passengers, but the Draft EIS didn't take the time to deal with that so I'm not going to either.


HUMAN-READABLE VERSION (Version 12/21/2008)

(c) 2008 by David B. Salzman, Ph.D.
The Bottom Line for MedInvest LRT v LowInvest BRT
0.53 lbs of CO2 emitted to move 1 passenger-mile in a car
39.76 lbs of CO2 emitted to move 1 TCar 1 mile
75.08 New passengers per 60-seat TCar for LRT to reach CO2 break-even
408% Avg 24x7 LRT occupancy needed to reach CO2 break-even
113,423,818 Gross lbs of CO2 emitted per year by LRT
73,049,938 Net lbs of CO2 emitted per year by LRT (corrected for displaced cars)
4.41 lbs of CO2 emitted to move 1 bus 1 mile
8.33 New passengers per 60-seat bus to reach CO2 break-even
13.9% Avg BRT occupancy needed to reach CO2 break-even
9,591,681 Gross lbs of CO2 emitted per year by BRT
-14,380,310 Net lbs of CO2 emitted per year by BRT (actually, savings)

THE DETAILS

Data Inputs: Comments Source
16.00 miles/route DEIS pg 6-3
60 seats/TCar DEIS pg 2-31
59 minutes/route for MedInvest LRT DEIS pg 6-3
3,372 TCarHrs/wk imputed from DEIS table 3-3
1738 trainHrs/wk imputed from DEIS table 3-3
550 Peak trainHrs/wk imputed from DEIS table 3-3
62,600 trips/wkdy DEIS pg 6-3
19,200 New trips/wkdy DEIS pg 6-3
$118.00 $/ton for coal in MD http://www.eia.doe.gov/cneaf/coal/quarterly/qcr_sum.html
$4.70 $/Mbtu for coal in MD http://www.eia.doe.gov/cneaf/electricity/epm/table4_10_a.html
210.20 CO2 lbs/Mbtu from coal http://www.eia.doe.gov/cneaf/coal/quarterly/co2_article/co2.html
33.1% Electricity Rcvd at TCar http://www.eia.doe.gov/emeu/mecs/mecs94/ei/elec.html
62,640 btu/mile to power each TCar http://cta.ornl.gov/data/tedb27/Edition27_Chapter02.pdf
22.4 Car MPG 2007 http://www.bts.gov/publications/national_transportation_statistics/html/table_04_23.html
19.33 CO2 lbs/gallon of gasoline http://www.epa.gov/otaq/climate/420f05001.htm
1.63 Avg # passengers/car http://www.bts.gov/publications/highlights_of_the_2001_national_household_travel_survey/html/table_a14.html
-186,400 VMT/wkdy for HiInvest LRT option DEIS pg 3-9
1,031 btu/cf CNG cf refers to cubic foot of (UNcompressed) natural gas at STP
0.122 CO2 lbs per cf CNG http://cdiac.ornl.gov/pns/faq.html
37,268 btu/mile of CNG BRT http://cta.ornl.gov/data/tedb27/Edition27_Chapter02.pdf
40,000 trips/wkdy for LowInvest BRT DEIS pg 6-3
11,400 NewTrips/wkdy for LowInvest BRT DEIS pg 3-5
7,967 BRT miles/wkdy DEIS pg 4-87

LRT TCars:
60 seats/TCar
0.98 hours/route
16.00 miles/route
3,372 TCarHrs/wk
1,738 trainHrs/wk
550 Peak trainHrs/wk
68.4% % TCars off-peak
1.94 Avg TCars/train
3,291,986 seat-miles/wk

LRT passengers:
62,600 trips/wkdy
19,200 New trips/wkdy
30.7% % New riders
7,825 trips/Sunday (guesstimate 1/8 of wkdy)
328,650 trips/wk
100,800 NewTrips/wk
8.92 Avg miles/trip (independent of # of passengers per car)
89.1% Avg occupancy (concocted to be ridiculously high)
53.46 Avg riders/TCar
75.08 | NewPassengers for CO2 break-even (in a 60-seat TCar?)
244.79 | | Total Passengers for break-even (impossible to save CO2 in any LRT)

At power plant:
$118.00 $/ton in MD
$4.70 $/Mbtu in MD
12,553 btu/lb coal
210.20 CO2 lbs/Mbtu
2.64 lbs CO2:coal

At MedInvest LRT train:
33.1% Electricity Rcvd
4,158 effective btu/lb
62,640 btu/mile
15.07 coal lbs/mile
39.76 CO2 lbs/mile
2,181,227 CO2 lbs/wk
0.74 CO2 lbs/TripMile
2.42 CO2 lbs/NewTripMile

Passenger car stats:
22.4 Car MPG 2007
22.4 Car MPG 2030 Obviously this will be a lot higher as cars improve and fleet mix changes!
19.33 CO2 lbs/gallon
0.86 CO2 lbs/mile
1.63 Avg Passengers/car
0.53 CO2 lbs/PassengerMile
4.73 net avg CO2 lbs/trip

Car miles avoided:
-186,400 VMT/wkdy HiInvest
-171,346 VMT/wkdy MedInvest
-21,418 VMT/Sunday (guesstimate 1/8 of wkdy)
-899,565 VMT/wk MedInvest
-40,159 gallons/wk MedInvest
-776,421 CO2 lbs/wk MedInvest

CO2 emissions by MedInvest LRT:
2,181,227 LRT CO2 lbs/wk emitted
-776,421 car CO2 lbs/wk displaced
35.6% net/gross CO2 (needs to be negative for any net savings)
1,404,807 net CO2 lbs/wk emitted
73,049,938 net CO2 lbs/yr (the LRT emits far more CO2 than the cars it displaces)
0.48 net avg CO2 lbs/TripMile
1.56 net avg CO2 lbs/NewTripMile

Compressed Natural Gas:
1,031 btu/cf CNG
0.122 CO2 lbs/cf CNG
118.33 CO2 lbs/Mbtu CNG
37,268 btu/mile CNG BRT
4.41 CO2 lbs/mile CNG BRT

LowInvest BRT:
40,000 trips/wkdy
11,400 NewTrips/wkdy
28.5% % New riders
5,000 trips/Sunday (guesstimate 1/8 of wkdy)
210,000 trips/wk
7,967 BRT miles/wkdy
41,827 BRT miles/wk

CO2 emissions by BRT:
184,455 BRT CO2 lbs/wk emiitted
59.4% Car derating for LowInvest LRT
-461,000 car CO2 lbs/wk displaced
-149.9% car CO2 displacement
-276,544 net CO2 lbs/wk
-14,380,310 net CO2 lbs/yr
-7.36 net avg CO2 lbs/NewTrip (BRT displaces CO2)
44.81 Avg riders/bus
-6.61 net CO2 lbs/mile BRT-cars (the BRT saves CO2)
8.33 | NewPassenger for CO2 break-even (9th new passenger saves CO2)
29.22 | | Total Passengers for break-even (30th total passenger saves CO2)


+ It's easy to put the machine-readable version of the spreadsheet into Excel. First, select and copy it from below and paste it into a word processor. Replace all instances of the five-letters "[tab]" with an actual tab character (e.g. the two characters "^t" in the Replace dialog box in Microsoft Word). Now select the corrected version from your word processor, copy it, bring up a fresh spreadsheet in Excel, click once on cell A10, and paste. If the cell stating "This should be in cell B10" is not in B10 on your computer, then figure out how far off you are, undo the paste, click on a cell that will put B10 where it was supposed to be, and try again. Then check that cell B148 is where it should be too. You can clean up the formatting (percentages, number of decimal places, etc.) by referring to the human-readable version.

SPREADSHEET-READABLE (Version 12/21/2008)

A10[tab]This should be cell B10[tab]
A11[tab](c) 2008 by David B. Salzman, Ph.D.[tab]
A12[tab][tab]
A13[tab]The Bottom Line for MedInvest LRT v LowInvest BRT [tab]
A14[tab]0.53[tab]lbs of CO2 emitted to move 1 passenger-mile in a car
A15[tab]39.76[tab]lbs of CO2 emitted to move 1 TCar 1 mile
A16[tab]75.08[tab]New passengers per 60-seat TCar for LRT to reach CO2 break-even
A17[tab]408%[tab]Avg 24x7 LRT occupancy needed to reach CO2 break-even
A18[tab]113,423,818[tab]Gross lbs of CO2 emitted per year by LRT
A19[tab]73,049,938[tab]Net lbs of CO2 emitted per year by LRT (corrected for displaced cars)
A20[tab]4.41[tab]lbs of CO2 emitted to move 1 bus 1 mile
A21[tab]8.33[tab]New passengers per 60-seat bus to reach CO2 break-even
A22[tab]13.9%[tab]Avg BRT occupancy needed to reach CO2 break-even
A23[tab]9,591,681[tab]Gross lbs of CO2 emitted per year by BRT
A24[tab]-14,380,310[tab]Net lbs of CO2 emitted per year by BRT (actually, savings)
A25[tab][tab]
A26[tab]THE DETAILS[tab]
A27[tab][tab]
A28[tab]Data Inputs:[tab]
A29[tab]16.00[tab] miles/route from DEIS pg 6-3
A30[tab]60[tab] seats/TCar from DEIS pg 2-31
A31[tab]59[tab] minutes/route for MedInvest LRT from DEIS pg 6-3
A32[tab]3372[tab] TCarHrs/wk from imputed from DEIS table 3-3
A33[tab]1738[tab] trainHrs/wk from imputed from DEIS table 3-3
A34[tab]550[tab] Peak trainHrs/wk from imputed from DEIS table 3-3
A35[tab]62600[tab] trips/wkdy from DEIS pg 6-3
A36[tab]19200[tab] New trips/wkdy from DEIS pg 6-3
A37[tab]$118.00[tab] $/ton for coal in MD from http://www.eia.doe.gov/cneaf/coal/quarterly/qcr_sum.html
A38[tab]$4.70[tab] $/Mbtu for coal in MD from http://www.eia.doe.gov/cneaf/electricity/epm/table4_10_a.html
A39[tab]210.20[tab] CO2 lbs/Mbtu from coal from http://www.eia.doe.gov/cneaf/coal/quarterly/co2_article/co2.html
A40[tab]33.1%[tab] Electricity Rcvd at TCar from http://www.eia.doe.gov/emeu/mecs/mecs94/ei/elec.html
A41[tab]62640[tab] btu/mile to power each TCar from http://cta.ornl.gov/data/tedb27/Edition27_Chapter02.pdf
A42[tab]22.4[tab] Car MPG 2007 from http://www.bts.gov/publications/national_transportation_statistics/html/table_04_23.html
A43[tab]19.33[tab] CO2 lbs/gallon of gasoline from http://www.epa.gov/otaq/climate/420f05001.htm
A44[tab]1.63[tab] Avg # passengers/car from http://www.bts.gov/publications/highlights_of_the_2001_national_household_travel_survey/html/table_a14.html
A45[tab]-186400[tab] VMT/wkdy for HiInvest LRT option from DEIS pg 3-9
A46[tab]1031[tab] btu/cf CNG from cf refers to cubic foot of (UNcompressed) natural gas at STP
A47[tab]0.122[tab] CO2 lbs per cf CNG from http://cdiac.ornl.gov/pns/faq.html
A48[tab]37268[tab] btu/mile of CNG BRT from http://cta.ornl.gov/data/tedb27/Edition27_Chapter02.pdf
A49[tab]40000[tab] trips/wkdy for LowInvest BRT from DEIS pg 6-3
A50[tab]11400[tab] NewTrips/wkdy for LowInvest BRT from DEIS pg 3-5
A51[tab]7967[tab] BRT miles/wkdy from DEIS pg 4-87
A52[tab][tab]
A53[tab]LRT TCars: [tab]
A54[tab]=B30[tab] seats/TCar
A55[tab]=B31/60[tab] hours/route
A56[tab]=B29[tab] miles/route
A57[tab]=B32[tab] TCarHrs/wk
A58[tab]=B33[tab] trainHrs/wk
A59[tab]=B34[tab] Peak trainHrs/wk
A60[tab]=1-B59/B58[tab] % TCars off-peak
A61[tab]=B57/B58[tab] Avg TCars/train
A62[tab]=B54*B56*B57/B55[tab] seat-miles/wk
A63[tab][tab]
A64[tab]LRT passengers:[tab]
A65[tab]=B35[tab] trips/wkdy
A66[tab]=B36[tab] New trips/wkdy
A67[tab]=B66/B65[tab] % New riders
A68[tab]=B65/8[tab] trips/Sunday (guesstimate 1/8 of wkdy)
A69[tab]=5*B65+2*B68[tab] trips/wk
A70[tab]=B67*B69[tab] NewTrips/wk
A71[tab]=(-B107/B70)[tab] Avg miles/trip (independent of # of passengers per car)
A72[tab]=B69*B71/B62[tab] Avg occupancy (concocted to be ridiculously high)
A73[tab]=B72*B54[tab] Avg riders/TCar
A74[tab]=B73*B91/B100[tab] | NewPassenger for CO2 break-even (in a 60-seat TCar?)
A75[tab]=B74/B67[tab] | | Total Passengers for break-even (impossible to save CO2 in any LRT)
A76[tab][tab]
A77[tab]At power plant:[tab]
A78[tab]=B37[tab] $/ton in MD
A79[tab]=B38[tab] $/Mbtu in MD
A80[tab]=(1000000*B78)/(2000*B79)[tab] btu/lb coal
A81[tab]=B39[tab] CO2 lbs/Mbtu
A82[tab]=B81*B80/1000000[tab] lbs CO2:coal
A83[tab][tab]
A84[tab]At MedInvest LRT train:[tab]
A85[tab]=B40[tab] Electricity Rcvd
A86[tab]=B80*B85[tab] effective btu/lb
A87[tab]=B41[tab] btu/mile
A88[tab]=B87/B86[tab] coal lbs/mile
A89[tab]=B88*B82[tab] CO2 lbs/mile
A90[tab]=B89*B56*B57/B55[tab] CO2 lbs/wk
A91[tab]=B89/B73[tab] CO2 lbs/TripMile
A92[tab]=B91/B67[tab] CO2 lbs/NewTripMile
A93[tab][tab]
A94[tab]Passenger car stats:[tab]
A95[tab]=B42[tab] Car MPG 2007
A96[tab]=B95*1.000[tab] Car MPG 2030 Obviously this will be a lot higher as cars improve and fleet mix changes!
A97[tab]=8.788*2.2[tab] CO2 lbs/gallon
A98[tab]=B97/B96[tab] CO2 lbs/mile
A99[tab]=B44[tab] Avg Passengers/car
A100[tab]=B98/B99[tab] CO2 lbs/PassengerMile
A101[tab]=B100*B71[tab] net avg CO2 lbs/trip
A102[tab][tab]
A103[tab]Car miles avoided:[tab]
A104[tab]=B45[tab] VMT/wkdy HiInvest
A105[tab]=B104*B65/68100[tab] VMT/wkdy MedInvest
A106[tab]=B105/8[tab] VMT/Sunday (guesstimate 1/8 of wkdy)
A107[tab]=5*B105+2*B106[tab] VMT/wk MedInvest
A108[tab]=B107/B96[tab] gallons/wk MedInvest
A109[tab]=B97*B108[tab] CO2 lbs/wk MedInvest
A110[tab][tab]
A111[tab]CO2 emissions by MedInvest LRT:[tab]
A112[tab]=B90[tab] LRT CO2 lbs/wk emitted
A113[tab]=B109[tab] car CO2 lbs/wk displaced
A114[tab]=1-B115/B112[tab] net/gross CO2 (needs to be negative for any net savings)
A115[tab]=B112+B113[tab] net CO2 lbs/wk emitted
A116[tab]=52*B115[tab] net CO2 lbs/yr (the LRT emits far more CO2 than the cars it displaces)
A117[tab]= B91*(1-B114)[tab] net avg CO2 lbs/TripMile
A118[tab]= B117*B92/B91[tab] net avg CO2 lbs/NewTripMile
A119[tab][tab]
A120[tab]Compressed Natural Gas:[tab]
A121[tab]=B46[tab] btu/cf CNG
A122[tab]=B47[tab] CO2 lbs/cf CNG
A123[tab]=1000000*B122/B121[tab] CO2 lbs/Mbtu CNG
A124[tab]=B48[tab] btu/mile CNG BRT
A125[tab]=B123*B124/1000000[tab] CO2 lbs/mile CNG BRT
A126[tab][tab]
A127[tab]LowInvest BRT:[tab]
A128[tab]=B49[tab] trips/wkdy
A129[tab]=B50[tab] NewTrips/wkdy
A130[tab]=B129/B128[tab] % New riders
A131[tab]=B128/8[tab] trips/Sunday (guesstimate 1/8 of wkdy)
A132[tab]=5*B128+2*B131[tab] trips/wk
A133[tab]=B51[tab] BRT miles/wkdy
A134[tab]=B133*B132/B128[tab] BRT miles/wk
A135[tab][tab]
A136[tab]CO2 emissions by BRT:[tab]
A137[tab]=B125*B134[tab] BRT CO2 lbs/wk emiitted
A138[tab]=B129/B66[tab] Car derating for LowInvest LRT
A139[tab]=B138*B109[tab] car CO2 lbs/wk displaced
A140[tab]=B141/B137[tab] net/gross CO2 (needs to be negative for any savings)
A141[tab]=B137+B139[tab] net CO2 lbs/wk
A142[tab]=52*B141[tab] net CO2 lbs/yr
A143[tab]=(B139/B88)/B86[tab] net avg CO2 lbs/NewTrip (BRT displaces CO2)
A144[tab]=B71*B132/B134[tab] Avg riders/bus
A145[tab]=B125*B140[tab] net CO2 lbs/mile BRT-cars (the BRT saves CO2)
A146[tab]=B125/B100[tab] | NewPassengers for CO2 break-even (9th new passenger saves CO2)
A147[tab]=B146/B130[tab] | | Total Passengers for break-even (30th total passenger saves CO2)
A148[tab]This should be cell B148[tab]