laboratory of ecosystem management life cycle environmental impacts of internet yves loerincik,...

Laboratory of ecosystem management

Life Cycle Environmental

Impacts of Internet

Yves Loerincik, Olivier Jolliet, Greg Norris


This study

1. The Internet infrastructure has been assessed in two cases:

- The Swiss Federal Institute of Technology (EPFL)

- The US

2. The Internet infrastructure has been assessed using:

- A Process LCA (PLCA)

- An Input-Output LCA (IO LCA) (A few slides on the theory on IO LCA are available at the end of the presentation for those interested)


Aims

1. To point out where are the main impacts related to the Internet infrastructure

2. To be able to take into account the Internet infrastructure when evaluating services like e-business, e-banking etc...

3. To compare the Process LCA (PLCA) and the Input-Output LCA (IO LCA)


The case study of the Swiss Federal

Institute of Technology (EPFL)

Internet infrastructure


Equipment Numberserver 68

switches 90router 22PC 6745

notebook 355printer 400cable 121225 meter

Step 1: Inventory of the necessary equipment for the Internet infrastructure at the EPFL

The functional unit = Internet infrastructure during one year


Corresponding annual costs (necessary to perform the IO LCA)

0

0.5

1

1.5

2

2.5

3

3.5

An

nu

al

co

st

[M$

]

5

Use

Production


0.00E+00

1.00E+07

2.00E+07

3.00E+07

4.00E+07

5.00E+07

6.00E+07

7.00E+07

8.00E+07

IO LCA Process LCA

En

erg

y [M

J]

papercableprinternotebookscreenrouterswitchcpuserver

Results: the non-renewable primary embodied energy (use phase and production)

cpu = control unit


Remarks

1. PCs are dominating (control unit (cpu) + screen)

2. Contribution of the switches and servers are significant

3. IO LCA value is two times the PLCA result


Non-renewable embodied primary energy, comparison between use and production phase

(relative value)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

serv

er I-

O

serv

er pr

oces

s

contr

ol un

it I-O

contr

ol un

it pro

cess

switc

h I-O

switc

h pro

cess

route

r I-O

route

r pro

cess

scre

en I-

O

scre

en pr

oces

s

note

book

I-O

note

book

proc

ess

print

er I-

O

print

er p

roce

ss

Use phaseProduction


Remarks

1. Use phase is in most of the cases dominating

2. The embodied energy during production is significant


0.00E+00

5.00E+06

1.00E+07

1.50E+07

2.00E+07

2.50E+07

3.00E+07

En

erg

y [M

J]

'

use

production

Non-renewable embodied primary energy, comparison between use and production phase

(absolute value)


Remarks

1. Use phase is more important in PLCA than in IO LCA (in relative values)

2. Results for the production is much higher with the IO LCA

Key parameter = the electricity and equipment prices (use to perform the IO LCA)


The case study of the United States Internet

infrastructure

Filippo Della Croce, Olivier Jolliet, Greg Norris, Yves Loerincik


Non-renewable primary embodied energy of the Internet infrastructure in place in the USA in 1999

using PLCA

0.00E+00

5.00E+10

1.00E+11

1.50E+11

2.00E+11

2.50E+11

Display

Deskt

op c

ompu

ter

Porta

ble c

ompu

ter

Serve

r

rout

ers

Mini

com

pute

r

Main

fram

e

Term

inal

Total

print

ers

Mod

ems

porta

ble p

hone

s

net s

cree

npho

nes

e-m

ail te

rmina

ls

web te

rmina

ls

net -

smar

t han

dlese

ts

LAN c

oppe

r

LAN fi

ber o

ptics

WAN c

oppe

r

WAN fi

ber o

ptics

Em

bo

die

d e

ne

rgy

[MJ

]


0.00E+00

5.00E+10

1.00E+11

1.50E+11

2.00E+11

2.50E+11

3.00E+11

3.50E+11

4.00E+11

Display

Deskt

op c

ompu

ter

Porta

ble c

ompu

ter

Serve

r

rout

ers

Mini

com

pute

r

Main

fram

e

Term

inal

Total

print

ers

En

erg

y [M

J]

Use

production

Non-renewable primary embodied energy and use phase energy of the Internet Infrastructure

in place in the USA in 1999


Comparison of various case study for the non-renewable primary embodied energy of computers

0

2000

4000

6000

8000

10000

12000

14000

MCC (PLCA) Miyamoto (PLCA) Atlantic Consulting(PLCA)

TU Munich (PLCA) IO LCA (LCANetbase)

En

erg

y [M

J]


0.0

100.0

200.0

300.0

400.0

500.0

600.0

700.0

Atlantic consulting (PLCA) TU Munich (PLCA) IO LCA

En

erg

y p

er k

g [

MJ/

kg]

Comparison of IO and process LCA for computers on a per kg basis


The higher value of the Input-Output LCA can be due to:

- Different energy efficiencies of US and Europe

- Different levels of comprehensiveness of the two methods

- Level of accuracy

Remarks

Laboratory of ecosystem management Different levels of comprehensiveness

In an IO LCA all the sectors of the economy are described, therefore all the

contributions are taken into account.For example (see next slide): the air

transportation, the hotels and lodging place and the automotive rental are normally not considered in a PLCA

Laboratory of ecosystem management Different levels of comprehensiveness

% to total upstream embodied

energy cumulative %Air transportation 12% 12%Electric services (utilities) 12% 24%Computer peripheral equipment 10% 34%Wholesale trade 10% 43%Semiconductors and related devices 9% 53%Petroleum refining 6% 59%Other electronic components 5% 64%Miscellaneous plastics products, n.e.c. 4% 68%Relays and industrial controls 3% 71%Gas production and distribution (utilities) 3% 74%Automotive rental and leasing, without drivers 2% 76%Hotels and lodging places 2% 78%Telephone and telegraph apparatus 2% 80%Aluminum rolling and drawing 2% 81%Motors and generators 1% 83%Blast furnaces and steel mills 1% 84%Sheet metal work 1% 85%Electron tubes 1% 85%Nonferrous wiredrawing and insulating 1% 86%Fabricated metal products, n.e.c. 1% 87%Legal services 1% 87%Paperboard containers and boxes 1% 88%Motor freight transportation and warehousing 1% 89%Power, distribution, and specialty transformers 1% 89%Gaskets, packing, and sealing devices 1% 90%Metal stampings, n.e.c. 1% 91%Real estate agents, managers, operators, and lessors 1% 91%Banking 1% 92%Eating and drinking places 1% 92%Management and consulting services, testing and research labs 1% 93%

Yellow = not normally considered in a PLCA


Difference in the energy efficiency of US and Europe

0.00E+00

2.00E+06

4.00E+06

6.00E+06

8.00E+06

1.00E+07

1.20E+07

Book

publi

shing

Bottle

d an

d ca

nned

sof

t drin

ks

Electro

nic c

ompu

ters

House

hold

vacu

um c

leane

rs

House

hold

audio

and

vide

o eq

uipm

ent

Prere

cord

ed re

cord

s an

d ta

pes

Drugs

Period

icals

En

erg

y [M

J/M

$]

Primary energy permillions $(consumer value)for the US

Primary energy permillion $ for theNetherland


Where does the difference between Process and IO-LCA come from

0

1000

2000

3000

4000

5000

6000

7000

8000

primary embodied

energy of a control unitfrom Input Output LCA

primary embodied

energy of a control unitfrom Input Output LCA

(corrected for Europe)

contributions to the

difference betweenprimary embodied

energy of a control unitfrom IO and process

LCA

primary embodied

energy of a control unitfrom process LCA

MJ

considered only by IO and not

by process LCA

luck of accuracy at the sector level in

IO LCA (+/-)

energy efficiency

improvements between 1992 and 1997 and

uncertainties in our calculations


Remark

We only took into account the necessary physical equipment of the Internet infrastructure. But, if we

have a closer look (see next slide), the main expenses related to a computer network come from

other sectors than equipment (for example software, maintenance, management, etc…).


Total Cost of IT Ownership, Forrester Consultants, 1997

Hardware26%

Training14%

Software14%

Management46%

Management Costs:

Administration 34%

End user downtime 35%

Co-worker time 14%

Application developm’t 14%

Disaster prevention 5%

Disaster recovery 12%

If the boundaries are extended


Annual Costs, EPFL Internet

0

1

2

3

4

5

6

7

cpu

scre

en

note

book

serv

er

switc

h

rout

er

print

erca

ble

telec

om

softw

are

traini

ng mgt

Ann

ual C

ost

s [M

$]

Use

Production


0.00E+00

2.00E+07

4.00E+07

6.00E+07

8.00E+07

1.00E+08

1.20E+08

1.40E+08

1.60E+08

IO LCA with TCO IO LCA

En

erg

y [M

J]

mgt

training

software

telecom

cable

printer

notebook

screen

router

switch

cpu

server

Considering the other expenses and using an IO LCA

TCO = Total Cost of Ownership


Future work

- Look at all the expenses related to the computer network (maintenance, software, buildings, security equipment, etc…) and

evaluate their impacts.

- Use an hybrid approach to assess in more details the important contributions (PC,

electricity, ?,…)


The next slides present very quickly the Input-Output theory used to

perform the Input-Output Life Cycle Assessment

Input-Output Theory


- Original work done by Wassily Leontief (1906-1999) in the 60 ’s.

- Economic theory/model to assess the economic system.

- Combined with environmental data (for example to perform LCA).

Input-Output Theory


1) The economy is divided in sectors.

The US economy is divided in around 500 sectors (Examples: poultry and eggs, cotton, crude petroleum and natural gas, cookies and crackers, cigars, newspaper, luggage, metal heat treating, water transportation, hospitals, child day care service, etc...)

2) Statistical data on the monetary flows between these sectors are collected.

Input-Output Theory


Tier 2 Tier ...

Sector 2

Sector 1

Sector 3

Sector …

Sector 2

Sector 1

Sector 3

Sector …

Sector 2

Sector 1

Sector 3

Sector …

Tier 0 Tier 1

Sector 1

Sector 2

Sector 3

Sector …

1 Million $

The structure of the economy

Input-Output Theory


Y)AI(X 1

The Input-Output equation

Y = The final demand (for instance 5 $ of aluminium)X = The necessary output of all the sectors to produce

the 5 $ of aluminium

Input-Output Theory


Y)AI(FXFf 1

f

= total impact

The environmental equation

X = total outputF = matrix of environmental impacts

Input-Output Theory

laboratory of ecosystem management life cycle environmental impacts of internet yves loerincik,...

Documents

internet infrastructure2

process lca plca

dell latitude cpi2

dell triniton tx7688321sun

main impacts

olivier jolliet

mmprizedell pe440036

epflthe functional unit