Estimation of ICT Sector footprints - ETSI docbox

Estimation of ICT Sector footprints Anders S.G. Andrae , Huawei & Peter M. Corcoran, National University of Ireland, Galway www.huawei.com

Author/ Email: Dr. Anders Andrae/[email protected] HUAWEI TECHNOLOGIES CO., LTD.

Contents

Case study: Electricity footprint of the ICT Sector 2012-2017

Experience of Huawei

Challenges related to estimating a Sector footprint

Possible ways to make the estimation

Electricity as indicator for ICT

The role of LCA in ICT footprint estimations

Situation: ICT electricity usage, footprint case study

Several investigatons of ICT Sector footprints, around eight

Explosive data traffic growth + Considerable efficiency gains

Common wisdom ”ICT Sector is ≈2% of global CO2e” is about to change in the coming years?

Target

https://www.researchgat e.net/publication/255923 829_Emerging_Trends_in _Electricity_Consumption _for_Consumer_ICT

Changing electricity share of ICT Sector from 2012 to 2017

Compared to GeSI SMARTer 2020

1) more scenarios 2) different scope 3) more transparency 4) more details on PCs 4) more details on TVs&Peripherals

Scenarios HUAWEI TECHNOLOGIES CO., LTD.

3

Worst, Best, and Expected are presented

Problems

What is the annual and global electricity footprint of the ICT Sector?

What is the share of ICT Sector of the total global electricity footprint?

How does the share change between 2012-2017?

Which Sectors of the ICT Sector are dominating the electricity footprint?

How do these Sectors change between 2012 and 2017?

HUAWEI TECHNOLOGIES CO., LTD.

4

SOLUTION

Previous work was studied and used as starting point. 2008-2012 is rather well studied and documented by at least 7 other groups.

A clear scope of the ICT Sector was set for Upstream(”LCA”)+Use:

Client Devices: Desktops, Monitors (Screens), Laptops, Smartphones, Tablets

TV Devices: TVs, Game Consoles, Set-top Boxes, A/V Receivers, DVD/BlueRays

Networks (CPE, Core Network, Office Networks, Networks between Data Centers, and WiFi Access, Wired Access, Wireless Access)

Data Centers (data processing, storage and HVAC infrastructures)

Data collected:

Electricity usage in manufacturing and use,

Shipped units, replacement rates, data traffic growth, energy saving and

growth. HUAWEI TECHNOLOGIES CO., LTD. Cut-off: network equipment in data centers

5

Devices

Device Use.

Several investigations were studied to find typical electricity usages

Different lifetimes were used to estimate the installed base

Improvements in energy saving was included

Device upstream

Used LCA studies and other reports to arrive at the electricity/unit

Improvement in energy saving year by year was included

Market reports gave the global annual shipped units


6

Device electricity in TWhrs in 2012


7

Devices: Expected Growth Scenario TV sets growth 2012-2017

Desktop PCs and monitors

growth rate CAGR 4.4% to around 1 billion

Installed bases of smartphones and tablets

Stagnating, (0.56 and 0.59 billion installed)

Installed base of laptops

Stagnating, 2.1 billion installed

grow 5% and 20% per year respectively

Annual efficiency improvements

for use electricity around 2% depending on the particular device category.

From 2008 on 5% annual improvement in manufacturing electricity per unit


8

Device electricity in TWhrs in 20122017 for Expected Growth Scenario


9

Networks

Network Use.

Includes”core” (core, metro, CPE, fixed) and mobile networks (RAN).

Started from others estimated (TWh) numbers from 2012

Model includes Mobile traffic Data (EB/year) and % carried by LTE

Model includes TWh/EB for mobile data and ”core” traffic

Model includes different improvements of the above TWh/EB

”Core” IP incl. Data centers network traffic grows from 2600 EB/yr to 8500 EB/yr

Mobile network traffic grows from 10.8 EB/yr to 134.4 EB/yr

Network upstream

Estimated from ”share of Use stage” as there were few reliable markets statistics (except for base stations) for Network Equipment


10

Networks: Expected Growth Scenario 20122017 – Use stage 85%

Data traffic

LTE CAGR 2012-2017 109%, Non-LTE 51%

LTE grows from 14% in 2012 to 45% in 2017 of Wireless Mobile Data Traffic

”Core” Network 14% growth rate for electricity from 2012 Energy efficiency “Wireless access” network LTE 0.73 TWhrs/EB and Non-LTE (0.73/1.46=) 0.5 TWhrs/EB.grows to contribute 8.1% of total Network electricity consumption ”Core” Network 0.14 TWhrs/EB in 2017 Wireless LTE and non-LTE efficiency

Core 26% CAGR of EB

5% annual TWh/EB improvement Core IP traffic: 10% year-on-year improvement in TWhrs/EB


11

Networks: Electricity in TWhrs in 20122017 for Expected Growth Scenario


12

Results&Conclusions: Annual footprint growth of ICT Sector 3,500 3,000

2,500

Best Case Expected

2,000

Worst Case

1,500

1,000 2012

2013

2014

2015

2016

2017

Best Case CAGR 1.8% which is below Global electricity CAGR of >2%

Worst Case CAGR 13.1% which is a doubling in 5 years

Expected Case is similar to Öko-Institute for EU27 from 2012-2020 HUAWEI TECHNOLOGIES CO., LTD.

13

Results&Conclusions: Share ICT Sector of Global electricity (with CAGR 3%) 15.0% 14.0% 13.0% 12.0% 11.0%

Best Case

10.0%

Expected Case Worst Case

9.0% 8.0% 7.0% 6.0% 2012

2013

2014

2015

2016

2017

Best Case decline from 7.8% to 7.4%

Expected Case increase from 7.8% to 9.4%

Worst Case increase from 7.8% to 12.5% HUAWEI TECHNOLOGIES CO., LTD.

14

Results&Conclusions: Change for Expected Case Scenario LCA, 16% Devices, 34%

LCA, 18% Data Centers, 15% Networks, 20%

Devices Devices, 47%

Networks Data Centers LCA

2012

Data Centers, 21%

Devices Networks Data Centers LCA

Networks, 29%

2017

The trend is that there is a very little difference between the ratios of these components regardless of the growth scenario. The combined contributions of networks & data centers will switch place with direct electricity usage of end-user devices HUAWEI TECHNOLOGIES CO., LTD.

15

Ways forward

Upstream facts

Identification/agreement of the major Equipment which are the ICT Sector

Agreed Minimum and Maximum value/piece for the Upstream ”indicator values” of these Equipment

Agree on the effect of including/excluding EoLT for the footprint in Sector calculations

Publications

Peer reviewed academic publications apparently need extensive review!


16

Thank you www.huawei.com

Copyright©2011 Huawei Technologies Co., Ltd. All Rights Reserved. The information in this document may contain predictive statements including, without limitation, statements regarding the future financial and operating results, future product portfolio, new technology, etc. There are a number of factors that could cause actual results and developments to differ materially from those expressed or implied in the predictive statements. Therefore, such information is provided for reference purpose only and constitutes neither an offer nor an acceptance. Huawei may change the information at any time without notice.

Experience Life cycle assessments performed › › › › › › ›

Radio Base Stations All sorts of mobile phones Tablets Metals FTTx Networks Radio Access Networks Cloud computing Networks

Sector analysis performed ICT Sector with a defined scope

Possible ways to make the estimation • Shipped units • Data traffic • Intensity measures and their change over time • Combine with LCA • Focus on electricity and energy usage

Electricity usage as an indicator for ICT Sector environmental impact •

Advantages › › › › › › › ›

•

Physical unit Statistics are abundants (IEA, etc.) High research interest for electricity markets LCA LCA studies often report electricity Most important energy form for ICT Energy drives most environmental impacts Climate Change can be readily estimated from electricity mixes (preferable marginal) Carbon taxes closely linked with electricity generation technologies

Disadvantages › ›

Just a metric and no actual impacts assessed Manufacturing and EoLT impacts are less driven by electricity production than other unit processes

The role of LCA in ICT Sector footprint estimations • • • • •

Raw Material Aquisition Production Transport Scrap generation Recycling

Challenges related to estimating Sector footprints

ICT Sector footprint estimations range from satisfyingly transparent (understandable which inputs give a certain output) to black box (not fully clear how the estimation was done)

• Lack of standardization The Scope of ICT Sector footprint studies have not been clearly defined.

• Lack of LCA studies of high quality For some important ICT Equipment LCA results are lacking Simplified approaches are likely enough for Sector footprint...but not preferable long-term.

Devices: Best Low Growth Scenario 2012-2017 • Installed base

› All kinds of devices remains stable

• Overall year-on-year improvement › use stage energy efficiency of 5% for most device categories › From 2008 on 5% annual improvement in manufacturing electricity per unit

Device electricity in TWhrs in 20122017 for Best Low Growth Scenario

Devices: (Worst) High Growth Scenario 2012-2017 • TV sets growth ›

•

Desktop PCs and monitors ›

•

growth rate CAGR 14% to around 1.58 billion

Installed bases of smartphones and tablets ›

•

growth of 2% per year, (0.64 and 0.67 billion installed)

Installed base of laptops ›

•

CAGR 5% to 2.6 billion installed

grow 15% and 25% per year respectively

Annual efficiency improvements

› ›

for use electricity vary from 1% to 5% depending on the particular device category. From 2008 on 5% annual improvement in manufacturing electricity per unit

Device electricity in TWhrs in 20122017 for High (Worst) Growth Scenario

•

Networks: Best Low Growth Scenario 20122017 – Use stage 90% Data traffic › › ›

•

•

LTE CAGR 2012-2017 109%, Non-LTE 51% LTE grows from 14% in 2012 to 45% in 2017 of Wireless Mobile Data Traffic Core 26% CAGR of EB

”Core” Network 7.5% growth rate for electricity from 2012 › LTE 0.5 TWhrs/EB and Non-LTE (0.5/1.46=) 0.34 TWhrs/EB. “Wireless access” network › ”Core” Network 0.135 TWhrs/EB grows to contribute 7.5% of total Wireless LTE and non-LTE efficiency Network electricity consumption › 5% annual TWh/EB improvement in 2017 Energy efficiency

›

Core IP traffic: 15% year-on-year improvement in TWhrs/EB

Networks: Electricity in TWhrs in 2012-2017 for Best Low Growth Scenario

•

Networks: (Worst) High Growth Scenario 2012-2017 – Use stage 80% Data traffic › › ›

•

•

LTE CAGR 2012-2017 109%, Non-LTE 51% LTE grows from 14% in 2012 to 45% in 2017 of Wireless Mobile Data Traffic Core 26% CAGR of EB

”Core” Network 20.4% growth rate for electricity from 2012 › LTE 1.37 TWhrs/EB and Non-LTE (1.37/1.46=) 0.94 TWhrs/EB. “Wireless access” network › ”Core” Network 0.142 TWhrs/EB grows to contribute 11.4% of Wireless LTE and non-LTE efficiency total Network electricity › 5% annual TWh/EB improvement consumption in 2017 Energy efficiency

›

Core IP traffic: 5% year-on-year improvement in TWhrs/EB

Networks: Electricity in TWhrs in 2012-2017 for Worst High Growth Scenario

Data Centers: Electricity in TWhrs in 2012-2017 for Best Low Growth Scenario

Data Centers: Electricity in TWhrs in 2012-2017 for Worst High Growth Scenario

Results: Change for Best Case Scenario LCA, 18% Data Centers, 15%


Networks, 20%

• 2012


LCA, 19%

Devices, 33%

Data Centers, 20%

Devices Networks Data Centers

Networks, 28%

LCA

2017

• In 2012 direct consumption by devices is just less than half of the total contribution The trend is that especially Networks Use and also Data Center Use are growing and Device Use is shrinking.

Results: Change for Worst Case Scenario LCA, 18% LCA, 18% Data Centers, 15% Networks, 20%

• 2012



Devices, 32%

Data Centers, 21%

Devices Networks Data Centers

Networks, 29%

2017

We can make a number of observations on this projected 2017 data: • direct consumption by devices is less than 1/3 of electricity; compare with 1/2 in 2012. • data centers + networks combined now represent 1/2 of electricity usage • LCA remains approximately at the same level of contribution

Data Centers

• Data Center Use. › › › ›

Scope: data processing, storage and HVAC infrastructures Network infrastructure within and between data centers is part of “Core” Network but was cut-off (Lambert p. 6(12)Sect.3.1) Started from Koomey’s 2010 baseline, extrapolated to 2012 Then Use the fixed rates (for electricity usage increase) determined from the previous “core” network data to model data center electricity growth rates: Low growth 7.5%, Expected growth 14%, High growth 20%

• Data Center upstream ›

Estimated from ”share of Use stage” as there were no reliable market statistics for any shipped Data Center Equipment

Data Centers: Electricity in TWhrs in 2012-2017 for Expected Growth Scenario

Results TWhrs: Comparision to SMARTer2020 in 2012 3500

2950

3000

•SMARTer2020 does not show manufacturing electricity details, but just CO2e/device/year

2500

2000

1812

1776

2012 SMARTER2020 1504

1500

1520

2012 Andrae&Corcoran BASELINE 2012 Mills, MAX

1000 704

863

2012 Mills, MIN

726

550 500

343

337 49

448 258

12

0

•PCs: SMARTer2020 has assumed larger installed base and use the same use stage consumption for laptops and desktops • Smartphones: SMARTER2020 added “ordinary” mobile phones to smartphones&tablets •Peripherals: Present study include TVs,STB,GC,A/V Reciever,DVD whereas SMARTER2020 include Monitor, Printer, STB, and Home router .

Results TWhrs: Comparision to SMARTer2020 in 2017(2020) 3500

3332.7

3000

2526.6 2500 2189

2000

2017 SMARTER2020 1985

2017 Andrae&Corcoran EXP 2017 Andrae&Corcoran BEST 2017 Andrae&Corcoran WORST

1500 1115.6 1000

840

811.7

662

612

562 600

414

500

343

792.1 600.9 480

585 446

353 269

90 62 48 72

102

37

37

99

0 PCs

Smartphones

Tablets

Peripherals

Networks

Data Centers

TOTALS