Co-Control of Urban Air Pollutants and Greenhouse Gases in Mexico City

J. Jason West, Patricia Osnaya, Israel Laguna, and Julia Martínez

Instituto Nacional de Ecología, México

with support from:

Integrated Environmental Strategies Program

US Environmental Protection Agency

National Renewable Energy Laboratory

Urban Joint Global

- Low-sulfur coal- Smokestack controls- Catalytic converters- Inspection and maintenance- Diesel particle traps- Evaporative controls

- Clean fuels: wood > coal > oil > gas > renewables- Energy efficiency- Carbon and energy taxes - Public transport and land use- Retirement of old vehicles- Efficiency standards for new vehicles

- Carbon sequestering- Forest management- Control of other GHGs (CH4, N2O, CFCs, SF6)- Geoengineering

Co-benefits and Co-control Studies

Control measures

Local emissions GHG emissions

Exposure and Concentrations

Health effects andEconomic benefits

What is the “co-benefit” for local air quality and for health from actions to control GHG emissions?

Co - benefits

Co-benefits and Co-control Studies

Control measures

Local emissions GHG emissions

Exposure and Concentrations

Health effects andEconomic benefits

Co - controlHow can we plan to achieve local and GHG objectives simultaneousely?

Goals of Co-control Study

“To support the capacity in Mexico to analyze and develop policies addressing local air pollution and climate change in an integrated manner.”

1) Unify diverse studies of measures for the control of local air pollution and of GHGs, into a harmonized database of options, which is consistent among measures.

2) Develop and apply quantitative methods of analysis of policies, based on linear programming (LP) and goal programming (GP), to analyze minimum cost programs that achieve objectives for multiple pollutants.

Mexico City

• Local air pollution dominated by transport emissions (80% NOX, 40% HC, 36% PM10).

• Overlapping environment/development goals:mobility, energy, poverty, air quality, climate.

Summary of costs and emissions reductions Measures applied locally

MeasuresCost (US$ million) Emissions reductions (ton/yr in 2010)

Public invest.

Private invest.

Total invest.

NPV (fuel)

PM10 SO2 CO NOX HC CO2

PROAIRE Total 6,529 7,740 14,269 4,913 5,180 591,206 121,096 99,907

PROAIRE – 22 measures in this study

6,330 7,740 14,070 4,887 972 590,972 115,622 99,880

This study – 22 measures from PROAIRE

9,934 13,025 22,959 7,656 3,767

15.0%

627

1.9%

1,138,167

50.1%

90,698

32.2%

137,259

23.1%

2,246,946

3.1%

This study – 22 measures from PROAIRE at their maximum levels

13,041 18,871 31,912 10,645 5,393

21.5%

796

2.4%

1,550,773

68.3%

120,106

42.6%

184,098

31.0%

3,267,473

4.6%

This study – GHG measures applied at their maximum levels

1,631 1,695 3,326 -714 321

1.3%

1

0.0%

2,670

0.0%

3,953

1.4%

19,232

3.2%

6,279,621

8.7%

This study – All measures applied at their maximum levels

14,671 20,556 35,237 9,931 5,714

22.8%

797

2.4%

1,553,443

68.4%

124,059

44.0%

203,330

34.2%

9,547,094

13.3%

Percents are with respect to total projected emissions in 2010.

Cost-effectiveness of CO2 and NOX

V1&2

V6

V8

V9

V12&13

V21

V22

V23

T25

T26

T27a

T27b

T28

T33

T35

T36

I2a

I2b

I2c

I7

S1

S4

LPG2

LPG3

LPG4

SOL1

SOL2

SOL3SOL4

HYB1

HYB2

HYB3

HYB4

G2

G3

G4

G5

G7

G11

G12

-5

0

5

10

15

20

25

30

35

0 10 20 30

Cost-effectiveness NOX

Co

st-

eff

ec

tiv

en

es

s C

O2

V1&2 V6

V8 V9

V12&13 V21

V22 V23

T25 T26

T27a T27b

T28 T33

T35 T36

I2a I2b

I2c I7

S1 S4

LPG2 LPG3

LPG4 SOL1

SOL2 SOL3

SOL4 HYB1

HYB2 HYB3

HYB4 G2

G3 G4

G5 G7

G11 G12

MoreLess

Le

ssM

ore

Linear Programming Formulation

Minimize: Cost = Σ AiCi

Changing: Activity levels of meausres (Ai)

Subject to restrictions:

1) Maximum levels, each measure: Ai (Ai)max

2) Minimum levels, each measure : Ai 0

3) Emissions reductions: Σ(AiEi,k) Tk

This can be a good tool when considering multiple pollutants simultaneously.

We developed this in Excel for easy application.

Minimize NPV (fuel), using PROAIRE Measures

PR

OA

IRE

0.0

0.5

1.0

1.5

2.0

2.5

3.0

V1&

2

V6

V8

V9

V12

&13

V21

V22

V23

T25

T26

T27

a

T27

b

T28

T33

T35

T36

I2a

I2b

I2c I7

S1

S4

PROAIRE measure

Act

ivit

y le

ve

l

Max. LevelLevel inPROAIRE

Min

. NP

V (

fuel

)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

V1&

2

V6

V8

V9

V12

&13

V21

V22

V23

T25

T26

T27

a

T27

b

T28

T33

T35

T36

I2a

I2b

I2c I7

S1

S4

PROAIRE measure

Act

ivit

y le

ve

l

Max. Level

Local PROAIRE Objectives,including other local measures

0

5,000

10,000

15,000

20,000

25,000

PROAIRE Min. NPV(fuel)

Min.Invest.Total

Min. NPV(fuel)

Min.Invest.Total

To

tal

Inv

es

tme

nt

(US

$ m

illi

on

)

GHG measures

Services

Industry

Transport

Other vehicles

Private vehiclesNPV (fuel)

PROAIRE measures only

Including other GHG measures

Local control with CO2 objectives

Minimize NPV (fuel) for PROAIRE objectives, and vary the restrictions for CO2 emissions. For all the local measures.

PROAIRE

Min. NPV

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

NP

V (

fue

l) (

US

$ m

ill.)

Minimize NPV (fuel)

PROAIRE

Min. NPV

17,000

18,000

19,000

20,000

21,000

22,000

23,000

24,000

25,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

To

tal I

nv

est

me

nt

(US

$ m

ill.)

Minimize NPV (fuel)

Local control with CO2 objectives

Minimize NPV (fuel) and total investment for PROAIRE objectives, and vary the restrictions for CO2 emissions. For all the local measures.

E

D

F Min. NPV

PROAIRE

AMin. Inv.

BC

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

NP

V (

fue

l) (

US

$ m

ill.)

Minimize Investment

Minimize NPV (fuel)E

FMin. NPV

PROAIRE

AMin. Inv.

B

C

17,000

18,000

19,000

20,000

21,000

22,000

23,000

24,000

25,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

To

tal I

nv

est

me

nt

(US

$ m

ill.)

Minimize Investment

Minimize NPV (fuel)

Local control with CO2 objectives

0

5,000

10,000

15,000

20,000

25,000

PR

OA

IRE A B C D E F

To

tal I

nve

stm

ent

(US

$ m

ill.)

GHG measures

Services

Industry

Transport

Other vehicles

Private vehicles

NPV (fuel)

Min. Investment

reduce CO2

Min. NPV (fuel)

reduce CO2

Local and CO2 control - including national measures

Minimize NPV (fuel) and total investment for PROAIRE objectives, and vary the restrictions for CO2 emissions.Including national measures.

D E

AMin. Inv.

PROAIRE

Min. NPVlocal

B C

17,000

18,000

19,000

20,000

21,000

22,000

23,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

To

tal I

nv

est

me

nt

(US

$ m

ill.)

Minimize Investment

Minimize NPV (fuel)E

D

A Min. INV

PROAIRE

Min. NPVlocal

B

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

NP

V (

fue

l) (

US

$ m

ill.)

Minimize Investment

Minimize NPV (fuel)

Local and CO2 control -including national measures

0

5,000

10,000

15,000

20,000

25,000

PR

OA

IRE A B C D E F

Tota

l In

vest

men

t (U

S$

mil

l.)

GHG national

GHG local

Services

Industry

Transport

Other vehicles

Private vehicles

NPV (fuel)

Min. Investment

reduce CO2

Min. NPV (fuel)

reduce CO2

Conclusions

For Mexico City – - PROAIRE has a significant global “co-benefit” (3.1%

of CO2).- Efficiency measures can reduce CO2 at a net cost-

savings, with high investment costs, and modest local emissions benefits.

- The benefits of simultaneously planning for local and global pollution are often small (but not zero).

For air quality / climate management –- A measure with good “co-benefits” may not be the best

way to solve problems simultaneously. It is important to include all possible measures in the analysis.

Acknowledgments

• CAM– V. H. Páramo, J. Sarmiento, R. Perrusquía, B. Valdez, M. Flores

– O. Vázquez, B. Gutiérrez, J. Escandón, O. Higuera

– C. Reyna, R. Reyes, S. Victoria

• INE – A. Fernández, V. Garibay, P. Franco, H. Martínez, A. García, A.

Guzmán, H. Wornschimmel

• US EPA and NREL– J. Renné, C. Green, D. Kline, J. Leggett, S. Laitner, S. Brant, K.

Sibold, L. Sperling, B. Hemming

• Others– M. Hojer, O. Masera, W. Vergara, R. Favela, J. Gasca, J.

Quintanilla, F. Manzini, A. Sierra, S. Connors, P. Amar

Contexts of Study

1) Local air quality management – PROAIRE and its reviews every two years.

2) Climate change – there is domestic and international interest in reducing GHG emissions in Mexico.

3) International Co-benefits research

Goals of Co-control Study“To support the capacity in Mexico to analyze and

develop policies addressing local air pollution and climate change in an integrated manner.”

1) Unify diverse studies of measures for the control of local air pollution and of GHGs, into a harmonized database of options, which is consistent among measures.

2) Develop and apply quantitative methods of analysis of policies, based on linear programming (LP) and goal programming (GP), to analyze minimum cost programs that achieve objectives for multiple pollutants:

- as a tool that CAM can use for informing decisions.- to explore the relationships between controls on local

pollutants and GHGs.- develop methods of analysis which are complementary to

Co-benefits methods.

Construction of a harmonized database of measures

** We conducted an open process, in which all of the offices of CAM participated.

Sources of data about the measures:- PROAIRE (2002-2010), and COMETRAVI (1999).- Studies of GHG measures at a national level (Sheinbaum,

1997; Sheinbaum y Masera, 2000).- Studies of other technologies (funded by World Bank):

- solar water heaters (Quintanilla et al., 2000).- reducing leaks of residential LPG (TUV Rheinland,

2000).- hybrid electric buses (World Bank report, 2000).

Construction of a harmonized database: Emissions and Costs

EMISSIONS –emissions reductions, with respect to the baseline, in 2010 (ton/yr), consistent with PROAIRE.

NOTE: it is not possible to compare our $/ton calculations with those in the literature, because we use emissions in 2010 only.

COSTS – PROAIRE reports undiscounted investment costs (public and private), while GHG studies present the discounted NPV (9% discount rate).

• It was not possible to estimate the NPV for all of the PROAIRE measures, with all of the changes in operation and maintenance expenditures.

We use investment costs and the NPV (fuel) as indicators.

Guide to the Harmonized Database

1&2) Public & private investment = sum of investments in capital from 2002 to 2010, without discounting.

3) Total investment = private + public.4) NPV (fuel) = costs of investment and expenditures for fuel and

electricity (2002-2010), and the salvage value in 2010. This is with respect to the baseline. Discounted to a NPV using a 9% discount rate. Does not include other social or environmental benefits.

5) NPV (all) = costs of investment and all of the operation and maintenance costs, with respect to the baseline, discounted to NPV.

6) Emissions reductions – in ton /yr in 2010.7) Maximum level – The maximum level of application of each

measure (the maximum feasible technically and practically), divided by the level in the Table.

– The level of application in PROAIRE is 1.0.

PROAIREDATABASE

Measures Public

Inv.Private

Inv.NPV PM10 SO2 CO NOX HC CO2

PROAIRE 89

(17)

THIS

STUDY

22

PROAIRE: 24 vehicle measures; 14 transport; 7 industry; 9 services; 15 conservation of natural resource; 8 health; 4 environmental education; 8 institutional strengthening.

THIS STUDY: 8 vehicle measures; 8 transport; 4 industry; 2 services

Information obtained in the document:

National Potential for the years 2000, 2005, 2010 in ton CO2 / yr

Implemented in 1997 -2010

Costs of reduction in US$ / ton CO2 (annualized)

(includes Investment, Operation and Maintenance)

Information required for the database:

Local fraction of application for CO2

Emissions of local pollutants (PM10, SO2, CO, NOx, HC)

Investment costs (million dollars)

NPV of each measure (US$/ton)

Mitigation measures for GHGs

a) Mexico City Metropolitan Area (MCMA)G2 Residential efficient lightingG3 Commercial efficient lightingG4 Pumping of potable waterG5 Electric motors in industryG7 Industrial cogenerationG11 Forest restorationG12 Agroforestry options

 b) Rest of the nation

GN2 Residential efficient lightingGN3 Commercial efficient lightingGN4 Pumping of potable waterGN5 Electric motors in industryGN7 Industrial cogenerationGN8 Wind electricity generationGN9 Temperate forest managementGN10 Tropical forest managementGN11 Forest restorationGN12 Agroforestry options

Electrical

Forestry

Electricity measures

  Assumptions of the effect of changes in electricity consumption on the generation within the MCMA

1 Completely outside of the MCMA

2 Entirely from plants within the MCMA

3 Considering the interconnected system

4 

Function of the relation consumption MCMA / generation MCMA

 %

0

100

3.1

20 

We consider scenario #3 to be most realistic.

Principal assumptions

• Our costs and emissions are correct – we are subject to the limitations of our data sources.

• We use the NPV (fuel) instead of the NPV (all), and our horizon is limited to 2010.

• It is possible to implement more or less of a measure (with respect to PROAIRE), with proportional costs and changes in emissions, until the maximum level.

• The measures are independent, and the costs and emissions are additive.

• The tons of each pollutant are equivalent.• These measures are all of the possible measures.• The analysis is static – it reflects decisions made today,

and do not reflect the ability to change decisions in time.

Minimize NPV (fuel), using PROAIRE Measures

PROAIRE Min. NPV (fuel)

Shadow prices

Public invest. 9,934 6,286

Private invest. 13,025 12,929

Total invest. 22,959 19,216

NPV (fuel) 7,656 6,168

PM10 3,767 4,355 0

SO2 627 627 3,019,951

CO 1,138,167 1,138,167 3,909

NOX 90,698 90,698 14,965

HC 137,259 137,259 20,021

CO2 2,246,946 2,614,201Costs are in US$ million, emissions in ton/yr in 2010, and shadow prices are in US$/(ton/yr).

What should be the local objectives?

HCs

PM10

Results from min. NPV (fuel)using all of thelocal measures.

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

120,000 140,000 160,000 180,000 200,000Reducción HC (ton/año 2010)

Co

sto

(E

UA

$ m

ill.)

Inversión total

VPN (comb.)

PROAIRE

A

C

PROAIRE

B

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

3,500 4,000 4,500 5,000 5,500Reducción PM10 (ton/año 2010)

Co

sto

(E

UA

$ m

ill.)

Inversión total

VPN (comb.)

PROAIRE

A

C

PROAIRE

B

What should be the local objectives?

CO

NOX

Results from min. NPV (fuel)using all of thelocal measures.

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

800,000 1,000,000 1,200,000 1,400,000 1,600,000Reducción CO (ton/año 2010)

Co

sto

(E

UA

$ m

ill.)

Inversión total

VPN (comb.)

PROAIRE

A

C

PROAIRE

B

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

60,000 70,000 80,000 90,000 100,000 110,000 120,000 130,000Reducción NOX (ton/año 2010)

Co

sto

(EU

A$

mill

.)

Inversión total

VPN (comb.)

PROAIRE

A

C

PROAIRE

B

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

800,000 1,000,000 1,200,000 1,400,000 1,600,000Reduccion en emisiones CO (ton/ano 2010)

Co

sto

(E

UA

$ m

ill.)

Inversion total

VPN (comb.)

PROAIRE

A

C

PROAIRE

B

Variation of costs with CO

Costs can be reducedwith a smaller reduction in CO, and with lessinvestment in privateauto measures.

0

5,000

10,000

15,000

20,000

25,000

30,000

PROAIRE A B C

Inv

ers

ión

To

tal (

EU

A$

mill

.)

Medidas GEI

Servicios

Industria

Transporte

Otros vehículos

Autos particulares

VPN (comb.)

Testable Hypothesis

• Emissions reductions targets for local air quality and global climate can be achieved less expensively if planned simultaneously, than if they were planned separately.

Cost (Urban + Global) <

Cost (Urban) + Cost (Global)

Testing the Testable Hypothesis

Indicator PROAIRE local targets

5 million tonnes/yr CO2

Local + global

Simultaneous

local/global

Total investment

17,815 1,157 18,972 18,321

NPV (fuel) 5,991 -1,248 4,743 5,280

PM10 4,212 7 4,219 4,203

CO 1,138,167 57 1,138,224 1,138,167

NOX 90,698 488 90,186 90,698

HC 137,259 2 137,261 137,259

CO2 2,459,519 5,000,000 7,459,519 5,000,000

Emissions reductions are tonnes per year in 2010. Costs are US$million.

Solutions when minimizing the total investment cost, using only local measures.

Goal Programming

• Alternative to linear programming– There can be many objectives (goals), with penalties if the

goals are not met.• Formulation:

Minimize sum of weighted deviations from goals:

Σ (dj+wj

+ + dj-wj

-)

where dj+ and dj

- are deviations from goals, and wj+ and wj

- are weights

Subject to restrictions:

1) Maximum levels, each measure: Ai (Ai)max

2) Minimum levels, each measure : Ai 0

GP example application

Indicator LP Solution

(min. invest.)

Goal Weight GP Solution

Total investment

18,585 16,736 1.0 20,675

NPV (fuel) 6,233 5,601 0.33 6,629

PM10 2,772 3,326 9.0 3,192

CO 1,115,703 1,115,703 0.005 1,032,456

NOX 86,665 103,998 0.02 95,442

HC 116,731 140,077 0.03 109,028

CO2 1,991,845 2,399,562

Emissions reductions are tonnes per year in 2010. Costs are US$million. Weights are $/$ or US$million / (tonne/yr in 2010).

Minimize NPV (fuel), now with the Metro (T25)

Min

. NP

V (

fuel

)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

V1&

2

V6

V8

V9

V12

&13

V21

V22

V23

T25

T26

T27

a

T27

b

T28

T33

T35

T36

I2a

I2b

I2c I7

S1

S4

Medida de PROAIRE

Niv

el d

e A

ctiv

idad

Wit

h M

etro

0.0

0.5

1.0

1.5

2.0

2.5

3.0

V1&

2

V6

V8

V9

V12

&13

V21

V22

V23

T25

T26

T27

a

T27

b

T28

T33

T35

T36

I2a

I2b

I2c I7

S1

S4

Medida de PROAIRE

Niv

el

de

Ac

tiv

ida

d

Minimize NPV (fuel), now with the Metro (T25)

PROAIRE Min. NPV (fuel)

Shadow prices

Min. NPV with Metro

Public invest. 9,934 6,286 8,837

Private invest. 13,025 12,929 12,949

Total invest. 22,959 19,216 21,786

NPV (fuel) 7,656 6,168 7,039

PM10 3,767 4,355 0 4,365

SO2 627 627 3,019,951 627

CO 1,138,167 1,138,167 3,909 1,138,167

NOX 90,698 90,698 14,965 90,698

HC 137,259 137,259 20,021 137,259

CO2 2,246,946 2,614,201 2,731,086Costs are in US$ million, emissions in ton/yr in 2010, and shadow prices are in US$/(ton/yr).

Local PROAIRE Objectives, using PROAIRE Measures

0

5,000

10,000

15,000

20,000

25,000

PROAIRE Min. VPN(comb.)

Min.Invers.Total

Min. VPN(comb.)

Min.Invers.Total

Inv

ers

ión

to

tal

(EU

A$

mil

lio

ne

s)

Servicios

Industria

Transporte

Otros vehículos

Autosparticulares

VPN (combustible)

Emisiones de PROAIRE

Emisiones PROAIRE

excepto SO 2

75% of Local PROAIRE Objectives

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

PROAIRE Min. VPN(comb.)

Min.Invers.Total

Min. VPN(comb.)

Min.Invers.Total

Inv

ers

ión

To

tal

(EU

A$

mil

lon

es

)

Medidas GEI

Servicios

Industria

Transporte

Otros vehículos

Autosparticulares

VPN (combutible)

Medidas de PROAIRE

Incluyendo otras Medidas GEI

Local and CO2 control -including national measures

Minimize NPV (fuel) for PROAIRE objectives, and vary the restrictions for CO2 emissions.Including national measures.

Min. NPVlocal

PROAIRE

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

NP

V (

fue

l) (

US

$ m

ill.)

Minimize NPV (fuel)Min. NPV

local

PROAIRE

17,000

18,000

19,000

20,000

21,000

22,000

23,000

0 2 4 6 8 10CO2 reduction (mill. tonnes CO2/yr 2010)

To

tal I

nv

est

me

nt

(US

$ m

ill.)

Minimize NPV (fuel)

Conclusions – Harmonized Database

1) PROAIRE measures can reduce emissions of CO2 in the MCMA by 3.1% in 2010.

- 50% CO2 from transport measures, and 50% vehicular.

- The costs increased and reductions in emissions changed significantly since PROAIRE.

2) The GHG measures can reduce emissions of CO2 by 8.7% in 2010, while their changes in local emissions are less (3.2% HCs, 1.4% NOX).

- This reflects that the majority of electricity is produced outside of the Metropolitan Area.

- Many of these measures have negative NPVs.

Conclusions – Application of the LP for Local Pollution

1) We develop the LP and GP as tools for planning to achieve multiple pollutant (local-global) co-control.

2) It is possible to achieve the local emission reduction goals in PROAIRE at less cost, changing the emphasis on measures.

- We estimate that the minimum cost can reduce by 20% (total investment and NPV (fuel)).

- Lower cost results are not possible because many PROAIRE measures are applied near their maximum levels.

3) Including other GHG measures, the NPV (fuel) can reduce significantly, with large reductions in CO2 emissions.

Conclusions – Management of Local Air Pollutants and GHGs

1) CO2 emissions can be reduced, with increases in the investment cost and decreases in the NPV (fuel), by applying GHG measures.

2) The benefits of simultaneously planning for local and global pollution are often small (but not zero).

3) Although a measure can have significant Co-benefits, other combinations of measures may be better to achieve local and global objectives.

- Measures which reduce CO2 outside of the metropolitan area should be considered also.