Fleet-Based LCA:
Comparative CO2 Emission Burden of
Aluminum and Steel Fleets
Randy Kirchain
Materials Systems Laboratory
Massachusetts Institute of Technology
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Life Cycle Analysis
Conventional Analysis Is Focused On A Single Product
Necessarily "Compresses" All Time-Dependent Emissions Into A
Single Value
May Be Appropriate In Some Cases
However, When Significant Emissions Differences Occur Over Long
Periods Of Time, May Provide An Incomplete Picture
Pertinent In The Case Of Automobiles, Where Use-Phase Emissions
Dominate LCA Inventories
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Study Objectives
Detailed treatment of temporal effects in life-cycle assesment
When do impacts occur?
When is "payback" achieved?
What if production is distributed over time?
Make analytical assumptions explicit
Reveal implications of analytical assumptions
Impacts
Impacts
Impacts
Massachusetts Institute of Technology
Cambridge, Massachusetts
Impacts
Materials Systems Laboratory
Temporal Effects on Life-Cycle Assessment
Conventional LCA studies have looked at impacts of single products
Production −> Use −> Disposal
Even for single products, effects are inherently distributed over time
Particulary true for long-lived products like automobiles
Typical Life Cycle Impacts
3
2
1
Use
Disposal
4
Production
Release of CO2
5
0
Years
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Issues with Conventional Analysis
120
Because impacts are distributed,
point estimates do not fully
describe a life cycle
Use
Production
100
A question arises:
When does the impact of one
product drop below that of
another?
80
Thousands
CO2 emissions (lbs)
Relative timing of impacts may be
important
60
40
20
0
Steel
Aluminum
Production emissions accounts for only BIW material.
Remainder of vehicle assumed to be same for both
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Timing of Life-Cycle Burden Can Be Important
120
Steel
lbs CO2
Thousands
100
Aluminum
Steel overtakes
Al in cumulative
CO2 burden
80
60
1
2
3
4
5
6
7
8
9
10
40
20
0
0
1
2
3
4
5
6
7
8
9
10
11
12
Years
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Once Recognized, Time Exposes Other Needs
Analysis Requirements:
Detailed treatment of time
New modeling approach
&RPSDULQJ
6LQJOH 9HKLFOH ,PSDFWV
OEV &2
Key Issues
Is a single product the correct
unit of comparison?
Production at one point in time
Production distributed over time
Does the performance of
production technologies remain
constant?
Do all units operate identically
over their life?
0
2
4
6
8
10 12 14 16 18
<HDUV
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Impact of Fleet Perspective
Single Vehicle:
Production at One Instance,
Use Distributed Over Time
Fleet of Vehicles:
Production
Massachusetts Institute of Technology
Cambridge, Massachusetts
AND
Use Distributed Over Time
Materials Systems Laboratory
Fleet Introduction Changes Rate of Life-Cycle Emission
200
Non-linear growth
results from
additions of vehicles
Thousand lbs CO2
150
100
New vehicles added at
regular intervals, each
contributes to fleet total
50
0
0
1
2
3
4
5
6
7
8
9
10
Years
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
&XPXODWLYH &2 GLIIHUHQFH
7KRXVDQGV OEV
Fleet Changes Shape and Timing of Recovery Curves
$O6WHHO
$O8/6$%
Crossover before 7 years for
single vehicle comparison
&XPXODWLYH &2 GLIIHUHQFH
WRQV
<HDUV
$O6WHHO
$O8/6$%
Crossover after 13 years
for fleet comparison
<HDUV
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Modeling Framework: Systems Dynamics
Explicit Consideration of Time, Rates of Production and Accumulation
Unit Of Analysis: Fleets In Production And Use
Fixed Production Rates
Use Begins Following Production
Retirement Follows US Fleet Patterns
End-of-Life Vehicle Materials Available For Reuse
Key Analysis Assumption:
"New" Vehicles Imply Incremental Increases In Production
Of Raw Materials, With Associated Incremental Effects
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
"Incremental" Production & Effects
Introduction Of Aluminum Vehicles Will Mean Large Change In
Consumption Patterns
"New" Aluminum Production Will Be Required
Analysis Examines Incremental Changes In Resource Use
Energy Consumption
Recycled Material Availability
Analysis Assumes That All Other Economic Factors and Demands Are
Unaffected By New Aluminum Consumption
Price Changes & Demand Sensitivity Not Assumed
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Baseline Simulation Framework
Start With A Zero-Car Fleet
Growing To Constant Fleet Size of ~150
Produce One "Car" Per Month
Steel Conventional
Aluminum Intensive (AIV)
ULSAB
Start Driving The Car Immediately Following Production
Retire Cars According To US Fleet Statistics
Collect ELV Materials For Reuse
Use Primary Material To Make Up Difference In Current Production
Calculate And Sum CO2 Production
Compare Totals For Each Fleet Alternative
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Base Case Assumptions
Distance driven per year:
Life of vehicle:
CO2/gal gasoline:
Fleet introduction:
Aluminum electricity source:
Secondary weight savings:
Fuel economy improvement:
Stamping Yeild:
11,400 miles (18,240 km)
average 12.2 years using fate model
22.9 lb CO2/gal
1 "vehicle unit"/month
5 years Coal, then marginal DOE grid
50%
5% per 10% weight reduction
50%
Steel
ULSAB
Aluminum
BIW weight - lb(kg)
816 (371)
612 (278)
444 (202)
Curb weight
3180 (1445)
2874 (1306)
2622 (1192)
Fuel economy 27.5 / 23.0 MPG 28.8 / 24.1 MPG
CO2/lb material
Massachusetts Institute of Technology
Cambridge, Massachusetts
1.24
1.24
29.9 / 25.0 MPG
19.4-12.6(primary)
1.0 (recycled)
Materials Systems Laboratory
Modeling Vehicle Fate
Single vehicle models
assume all vehicles retire
at same time
All vehicles lasted
10-12 years
Cumulative Retires
Schematic Vehicle Retirement
Dynamic model can use
real data about vehicle
retirement
Some vehicles crash
and retire quickly
Others last a long time
100%
80%
60%
40%
20%
0%
1
3
2
5
4
7
6
9
8
10
11 13 15 17 19
12 14 16 18 20
Vehicle Age
Years
Base case scenario
generates average vehicle
life of 12.2 years
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Implications Of Vehicle Fate
Early Vehicle Retirement
Has Two Effects
Recycled Content of AIVs
Massachusetts Institute of Technology
Cambridge, Massachusetts
50%
25%
239
205
171
137
103
69
0%
35
In early years, less
vehicles on the road
This means less miles
driven
Vehicle Fate
Model
75%
1
Scrap aluminum
becomes available
more quickly
This reduces total
primary aluminum
usage
Recycled Al Content
100%
Months
Materials Systems Laboratory
Electrical Power Source Assumption
"5 years Coal, then marginal DOE grid"
1. Current non-fossil production part of base load generation
2. "New" aluminum production will require non-base loads
Less efficient, more polluting - currently coal-fired facilities
3. Eventually, utilities will build new base load capacity to satisfy
increases in demand
Forecasts predict combined cycle natural gas generation
"New" hydropower currently infeasible, although sites exist
Regulatory hurdles
Environmental concerns
Competing uses (recreation, etc.)
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Base Case Results: Years to Recover CO2 Burden
Al - Steel (23.5mpg)
300
Al - Steel (27mpg)
Al - ULSAB (27mpg)
200
tons
Cumulative CO2 difference
400
100
0
-100
-200
-300
10
80
150
220
290
360
Months
Base Case Crossover Points
Time Until Aluminum Fleets to Produce Less CO2 than
Conventional Steel: 11.8 - 13.5 years (23.5 - 27 mpg)
ULSAB:
23.7 - 27.7 years
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Robustness of Result
Crossover points are sensitive to some initial assumptions
Base fuel economy
Source of electricity
Vehicle weight savings
Impact of Base Fuel Economy
Years to Crossover
CO2 Burden Crossover
16
15
14
13
12
11
23
27.5
32
Vehicle Fuel Economy
Miles per Gallon
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Role of Aluminum Electricity Source
Results are sensitive to assumptions about Al electricity source
Conventional LCA relies on currently used resources
Current electrical grid
Industry reported electrical sources
Because manufacturing aluminum autos represents a significant
increase in aluminum production, we looked at marginal effects
Expansion in electrical demand
Current Al electricity sources are fully utilized
DOE forecast for expansion
Short term: Use of current facilities
Mid term: Largely new natural gas facilities
(12% coal, 3% renewables)
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Sensitivity To Aluminum Electricity Source
Nevertheless, other assumptions may be credible
Evaluating on marginal burden discourages investment in clean
energy sources
Other groups may choose most favorable scenario
Impact of Source of Aluminum Electricity
CO2 Burden Crossovers - (23.0 mpg / 27.5 mpg basis)
Source of Electricity
Coal - 5 years,
then Marginal Grid
Coal - 10 years,
then Marginal Grid
Current Grid
IPAI Mix
Al
vs.
Conventional Steel
Al
vs.
ULSAB
11.8 / 13.5
23.7 / 27.6
13.6 / 15.3
36.9 / 31.7
9.8 / 11.4
5.3 / 6.2
* Crossovers in years
20.9 / 24.3
13.3 / 15.4
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Robustness of Result
Potential fuel economy improvement can effect result
Aggressive secondary weight savings
Strong response of fuel economy to weight
Some scenarios push crossover to approximately average vehicle life
Fuel Economy
Improvement
per Mass Reduction
Secondary
Weight
Savings
5% : 10%
5% : 8%
50%
13.5
11.6
100%
11.1
9.6
* Crossovers in years
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Conclusions
Base case analysis shows a substantial recovery period before CO2
benefits of aluminum vehicles are realized
Al vs. Conventional Steel:
Al vs. ULSAB:
11.8 - 13.5 years (23.5 - 27 mpg)
23.7 - 27.7 years
Results are sensitive to analytical assumptions
Fuel economy improvement
Electricity source
Credible set of assumptions can result in simulated recovery periods
less than or equal to current average vehicle lifetime
Results may differ for emissions other than CO2
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory
Future Work
Continuing Validation and Testing of Simulation Assumptions
Electrical Power Source Scenarios
Energy In Post-Smelter/Furnace Processes (i.e., Rolling)
Influences of Changes in Key Rates
Vehicle Production/Introduction
Size of Steady-State Fleet
Rate of Vehicle Use
Displacement of an Existing Fleet
Alternative Recycling Accounting Methods
Emissions in Addition to CO2
Massachusetts Institute of Technology
Cambridge, Massachusetts
Materials Systems Laboratory