Download Engineering Systems: Broadening Engineering Research

Flexible Global
Climate Change Policy
ESD.71 Application Portfolio: December 2009
Nidhi R. Santen
Ph.D. Student, 2nd Year
Engineering Systems Division
[email protected]
© 2009 Nidhi R. Santen
Presentation Outline








Problem Overview (1)
Application System (1)
Sources of Uncertainty (3)
Fixed and Flexible Designs (1)
Decision Tree Analysis (2)
Lattice Analysis (2)
Conclusions (2)
Questions?
© 2009 Nidhi R. Santen
Problem Overview
 Current climate change policy decisions are based on
deterministic views of the future events, dominated
by considerations of a single optimal carbon emissions
path.
 Many aspects of the future remain highly uncertain,
creating the need for flexible climate policies.
 Implementing an R&D-inducing carbon tax policy
today provides a form of “insurance” against future
carbon-emissions related climate damages,
representing a important form of flexibility.
 The following project examines the opportunities that
flexible global climate change policy can have on the
overall net present welfare of the global economy.
© 2009 Nidhi R. Santen
Application System
 Key Components of System
 Aggregated Global Economy, includes
 Production = f(Capital, Labor, Fossil-Fuel
Energy Sector)
 Physical Environment
 Population’s Utility Function (Preferences for
Consumption v. Investment)
 Dynamic Integrated model for Climate
and the Economy (DICE-99) is used as
the evaluation model for system
performance
 Timeframe is 2015 through 2335
© 2009 Nidhi R. Santen
Sources of Uncertainty (1 of 3)
Total Factor Productivity Growth Rate:
TPF is the contribution to economic output not accounted for by inputs such as labor and
capital; level of technology in the economy
Histogram: Total Factor Productivity
60
40
30
20
10
0
16.00%
16.06%
16.11%
16.17%
16.23%
16.29%
16.34%
16.40%
16.46%
16.51%
16.57%
16.63%
16.69%
16.74%
16.80%
16.86%
16.91%
16.97%
17.03%
17.09%
17.14%
Percent Per Decade
50
Frequency
Data: MIT Joint Program
© 2009 Nidhi R. Santen
Sources of Uncertainty (2 of 3)
Emissions Intensity Growth Rate:
Emissions intensity is the trend in CO2-equivalent emissions per unit of output without a
carbon-reducing policy in place
Histogram: Growth Rate of Sigma
50
45
40
30
25
20
15
10
5
0
-0.09
-1.08
-2.08
-3.07
-4.07
-5.07
-6.06
-7.06
-8.06
-9.05
-10.05
-11.04
-12.04
-13.04
-14.03
-15.03
-16.02
-17.02
-18.02
-19.01
-20.01
Frequency
35
Percent Per Decade
Data: MIT Joint Program
© 2009 Nidhi R. Santen
Sources of Uncertainty (3 of 3)
Climate Feedback
A cloud-related parameter that represents the sensitivity of the climate to GHGs
Histogram: Climate Feedback
600
Frequency
500
400
300
200
100
0
1.65
4.10
6.54
8.98
11.42
Climate Feedback
Data: MIT Joint Program
© 2009 Nidhi R. Santen
Fixed and Flexible Designs
 Fixed Design
 In both studies: Business-as-usual case with
no carbon-tax policy
 Flexible Designs
 Investigation 1 (2 Period): High or low
carbon tax policy implemented in Period 1
with an option to change tax level in Period
2.
 Investigation 2 (6 Period): Option to
implement a medium carbon tax at any
period.
© 2009 Nidhi R. Santen
Decision Analysis (1 of 2)
Policy Design Alternatives and Uncertainties
Design
Fixed “No” Policy (“Business-as-Usual”)
Flexible Policies (Carbon Taxes)
Uncertain Parameter
Emission Intensity
Growth Rate (σt)
Climate Feedback
Parameter (λt)
Emissions Reduction (µt) ($ Tax)
No Control ($0 per ton) Both Decision
Points
2015: High ($30) / 2065: High ($80)
2015: High ($30) / 2065: Low ($30)
2015: Low ($10) / 2065: High ($80)
2015: Low ($10) / 2065: Low ($30)
High
Medium
Low
P(σ=-30.055) = 0.185
P(σ=--15.885) = 0.63
P(σ=-1.082) = 0.185
P(λ=4.682) = 0.185
P(λ=2.908) = 0.63
P(λ=1.134) = 0.185
Decision Analysis Components
Stages: 2 (2015-2055 and 2065-2335)
Decision in Period 1: Tax High, Tax Low, or No Tax
Uncertainties Considered: Emission Intensity Growth Rate and Climate Feedback
Payoff Value: NPV Welfare
© 2009 Nidhi R. Santen
Decision Analysis (2 of 2)
Decision Tree Solution Optimal Strategy:
Tax Low in Period 1; Tax High in Period 2
Decision Tree Analysis VARG Curve
1
0.9
0.8
Probability
0.7
0.6
0.5
0.4
Risk Assumption
Preferred Period 1 Design
0.3
P10
Tax Low
0.2
P15
Tax High
0.1
P50
No Tax (Inflexible Case)
P75
No Tax (Inflexible Case)
P90
No Tax (Inflexible Case)
0
966.6
966.8
967
967.2
967.4
967.6
967.8
968
NPV
Tax High
Tax Low
968.2
968.4
No Tax (No Flexibility-Fixed)
Decision Tree Analysis VARG Curve
© 2009 Nidhi R. Santen
Lattice Analysis (1 of 2)
Policy Design Alternatives




BAU No-Tax Case
Option to Begin Implementing a $45 per ton Carbon Tax in any
Period (Modeled as a “Call Option”)
Uncertainty Considered: TPF Growth Rate
“How long should we wait to implement a carbon tax?”
Probability Density Function
35.0%
30.0%
Probability

25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
4.1797
4.0374
3.8999
3.7671
3.6388
3.5149
3.3952
Total Factor Productivity Growth Rate (%)
© 2009 Nidhi R. Santen
Lattice Analysis (2 of 2)
Dynamic Programming Decision Analysis Optimal Strategy:
Always Decide to Implement $45 per ton Carbon Tax
t=0
NO
t=1
YES
t=2
YES
t=3
YES
t=4
YES
t=5
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
t=6
VARG curves - Carbon Tax Implementation Option
100.0%
90.0%
80.0%
Risk Assumption
Preferred Period 1 Design
P10
Flexible Case
P25
Flexible Case
P50
Flexible Case
P75
Flexible Case
P95
Flexible Case
Cumulative Probability
Period
Exercise
CALL
OPTION?
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
NPV ($)
Flexible
Inflexible
© 2009 Nidhi R. Santen
Conclusions
 Investigation 1 (Decision Tree Analysis):
The value of flexibility was $0.019316
trillion for the optimal strategy.
 Investigation 2 (Lattice Analysis): the
value of the call option to implement a
carbon-tax when deemed appropriate
was $0.51 trillion.
 Flexible policy strategies are chosen
over inflexible policy strategies in both
investigations.
© 2009 Nidhi R. Santen
Conclusions
Total Carbon Emissions by Strategy
20
18
16
14
GtC
12
10
8
6
4
2
0
Year
No Tax (Inflexible Policy)
Decision Tree Analysis Optimal Strategy
Lattice Decision Analysis Optimal Strategy
© 2009 Nidhi R. Santen
Conclusions
Atmospheric Temperature Increases by Strategy
5
Degrees Celsius Above 1900
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
Year
No Tax (Inflexible Policy)
Decision Tree Analysis Optimal Strategy
Lattice Decision Analysis Optimal Strategy
© 2009 Nidhi R. Santen
Thank You!
Questions?
[email protected]
Acknowledgments
Professor Richard de Neufville
TA Michel-Alexandre Cardin
© 2009 Nidhi R. Santen