Energy Consumption

  • GRI 302
  • GRI 302-1
  • GRI 302-2
  • GRI 302-3
  • GRI 302-4
  • GRI 305
  • GRI 305-1
  • GRI 305-2
  • GRI 305-3

Our network infrastructure and the server capacity provided by our data centers are needed to supply internet and telecommunications services to our customers. At the same time, their operation accounts for most of our emissions. Roughly half of the Group’s expenditure on electricity comes from our global data center operations. Most of the other half results from our fiber-optic network operations; the small remainder is attributable to our office buildings. Further details on the information given in the following table can be found in the relevant sections.

Energy consumption(1)

2018

2019(2)

2020(3)

Electricity consumption – data centers in MWh

149,012.53

133,807.96

131,025.85

Electricity consumption – fiber-optic network in MWh

60,113.43

59,599.17

56,665.61

Electricity consumption – office buildings in MWh

7,430.57

7,068.83

6,556.84

Total electricity consumption in MWh

216,556.53

200,475.96

194,248.29

Heating consumption (natural gas and heating oil) in MWh

2,886.65

2,859.01

3,130.86

Gasoline consumption in MWh

285.34

607.06

756.53

Diesel consumption in MWh

14,624.40

13,223.90

8,824.18

Total fuel consumption in MWh(4)

14,909.74

13,830.96

9,580.71

Total energy consumption in MWh

234,352.93

217,165.94

206,959.87

Revenue in € million

5,102.9

5,194.1

5,366.2

Energy intensity (ratio of energy consumption to revenue) in Wh/€

45.93

42.07

38.57

(1) Data capture was expanded and enhanced in many areas in 2018 and further companies were included.

(2) The figures for 2019 were adjusted.

(3) The decline in diesel consumption and total consumption in 2020 is due to the restrictions caused by the COVID-19 pandemic.

(4) Conversion factors/calorific values: gasoline: 8.5 kWh/l; diesel: 9.7 kWh/l.

Electricity Consumption in Our Data Centers

1&1 IONOS Holding’s high-performance data center operations are geo-redundant, making the services we provide to our customers as secure and reliable as possible. This approach ensures that application operations are not restricted by power outages or environmental influences, and gives our customers unconstrained, permanent access to them. This added security comes at the cost of increased electricity consumption, which we combat by using renewable energy and constantly improving our energy efficiency.

Climate Neutrality and the Use of Renewable Energy

We have used electricity from renewable energy sources for our data centers in Germany for more than a decade. In the meantime, green electricity is the main source of supply for our locations in all countries. We now only still use certificates to offset carbon emissions at a few locations. This allows us to operate 1&1 IONOS Holding’s data centers climate-neutrally. 1&1 Mail & Media Applications SE’s servers are all located in Germany at IONOS data centers, meaning that these also benefit from the latter’s climate-neutral operations, like IONOS’s own servers. Using renewable energy from nearby geographical areas – mostly in the same country or an adjacent region – is another priority.

The following table shows the electricity consumed in our data centers and other data centers belonging to the Group, along with the carbon emissions we avoided.

Electricity consumption and carbon emissions,
data centers(1)

2018

2019(2)(3)

2020(2)

Electricity consumption in MWh

149,012.53

133,807.96

131,025.85

Energy intensity (ratio of electricity consumption to revenue) in Wh/€

29.20

25.76

24.42

CO2 equivalents in tonnes(4) (Scope 2)

0

0

0

Avoided CO2 equivalents in tonnes(5)

65,642.00

57,828.26

48,891.36

(1) The figures refer to our 10 own and our rented data centers; from 2018 onwards, they relate to the data centers for IONOS, Arsys, Fasthosts, and Strato, plus the data centers for 1&1 IONOS Cloud, home.pl, InterNetX, and World4You. Electricity consumption for 1&1 Drillisch’s data centers is included under the electricity consumption data for its office buildings (see page 85), since it is not currently possible to capture consumption separately.

(2) The reduction in electricity consumption in 2019 and in the 2020 reporting period is due to a number of factors; see the section entitled “Increased energy efficiency” on page 82.

(3) The 2019 figures for electricity consumption, energy intensity, and avoided CO2 equivalents were adjusted retrospectively.

(4) Climate-neutral thanks to the mix of renewable energy sources and carbon offsets. The figures for CO2 equivalents show the contribution made to the greenhouse gas effect. In addition to CO2 itself, they comprise other greenhouse gases such as methane and dinitrogen monoxide.

(5) Approximate figure calculated on the basis of the average annual and country-specific CO2 emissions per kilowatt-hour of electricity consumed; emissions from the electricity mix in the various countries are falling over time, due among other things to the growing proportion of renewable energy used.

Managing Consumption with an ISO 50001-compliant Energy Management System

In 2018, we took the decision to introduce an energy management system (EnMS) complying with ISO 50001. This EnMS covers the IONOS and Strato data centers that we operate ourselves and in which we can therefore manage energy consumption. Using it allows us to continuously pursue our goals of increasing energy efficiency and ensuring transparency. As a result, the EnMS helps identify and stop potential energy wastage, hence cutting costs, and to identify and comply with external requirements such as regular energy audits. In this way, it makes an important contribution to sustainability management and can have a positive impact on our reputation and market position. The EnMS was successfully audited and certified in the 2020 reporting period.

The head of TechOps Hosting at IONOS, and the Data Center unit manager at Strato are responsible for the EnMS and its strategic focus. In addition, a designated energy manager ensures that the EnMS is aligned with our targets, and manages the Energy Team. The latter comprises the regional staff responsible for this topic at the IONOS and Strato data centers in the various countries concerned (Germany, France, Spain, the United Kingdom, and the USA). In addition, a management system manager has been appointed.

Increased Energy Efficiency

The air conditioning system at the Baden Airpark data center was successfully renewed in 2020. As a result, we were already able to make energy savings of 15% in the reporting period. 2021 will see the start of construction of a new, high-performance data center in the Birmingham region of the United Kingdom. This will replace the existing UK data center in Gloucester in coming years. The project aims both to increase capacity and to enhance operational efficiency in the United Kingdom. Current plans are for an improvement in energy efficiency of at least 20% in comparison to the existing data center.

In addition, we are implementing a continuous stream of projects and measures to reduce electricity consumption and increase energy efficiency:

  • We are steadily increasing capacity utilization per server, thus reducing the number required.

  • We replace old servers with more modern and more energy-efficient hardware in good time.

  • Some server hardware is built to order for United Internet, allowing us to avoid unnecessary components and deploy, among other things, energy-saving processors and power supply units that are designed to minimize heat losses.

  • The web hosting system used by United Internet Group companies is a highly optimized, proprietary, Linux-based system that allows data from thousands of customers to be managed on a single server, and hence optimizes our use of resources.

  • Virtualization is increasingly enabling us to replace bare metal servers by virtual servers.

  • The use of containers avoids the need for redundant operating system kernel operations; the kernel is shared by all instances, enabling even more elastic, load-based scaling of the IT resources provided.

First Experiences of Using Machine Learning for Optimizing Server Cooling

Another project involves analyzing data and using machine learning algorithms to predict customers’ storage requirements. This approach can improve the service we provide to our customers even further, make more efficient use of resources, and hence reduce energy and hardware requirements. Our data scientists also continuously pilot other potential applications for data analysis and AI.

Energy audits can also identify levers for improving energy efficiency and reducing the Group’s energy costs. The German Energy Services Act (“Energiedienstleistungsgesetz” – EDL-G) requires energy audits to be performed by an independent auditor every four years.

Balanced Design – The New Standard for All Future Data Center Projects

The demands placed on data centers are increasing all the time – both in terms of performance and regarding environmental aspects. The challenge for operators is to find a balance between ensuring high-availability, secure operations on the one hand, and cost and energy efficiency on the other. When it became clear that capacity requirements at one of the Group’s data centers in Spain would be exceeded, expanding our own data center soon emerged as a cost-effective solution. The next step was to design this so as to offer the right combination of availability, energy efficiency, construction time, and simple, flexible operations. Structured project planning allowed IONOS to implement an extremely cost-effective data center. The new design focuses on the elegant arrangement of the technical components and is based on the concept of distributed redundancy. In addition, the data center features adiabatic free cooling, in which the evaporation chill produced by air and water is used for cooling.

The new data center design also features a “pay as you grow” approach. The center has been constructed in such a way that it can be extended on the fly at any time if needed. Only essential components are installed and operated to start with – a highly cost-effective approach. As a result, there are no excess units that have to be replaced at some point because they are showing signs of age even though they have never been used. This conserves valuable resources and saves on maintenance costs, while equipment that has not yet been installed also does not consume any power. IONOS uses this approach for large components such as UPS units, cooling generators, and cooling units. The design of the new data center gives IONOS a scalable, energy-efficient, state-of-the-art server and cloud infrastructure that offers high availability and an attractive price-performance ratio to match. The success of the design means it will be used as the new standard for all future data center projects at IONOS.

Electricity Consumption by Our Fiber-optic Network

Our fiber-optic network ensures economic efficiency and enables society to benefit from increasingly powerful, rapid data transfer. Today’s technology permits internet speeds of up to 100 GBit/s and represents the most powerful transmission technology by far. What is more, demand for higher bandwidth is constantly rising. However, since fiber optics are not available everywhere in Germany yet, 1&1 Versatel is continuing to constantly expand our fiber-optic network. In 2019, extensive additions were made to 1&1 Versatel’s network, which was further optimized for our customers using more powerful network technology. At the end of the 2020 reporting period, our fiber-optic network was approximately 50,900 km long, compared to roughly 48,500 km the previous year. The network has been continuously expanded in recent years and is one of the largest and most powerful in Germany, already providing a large number of cities and municipalities with high-speed internet services.

Since 2012, 1&1 Versatel has managed energy consumption using an ISO 14001-based environmental management system (EMS). We also give reducing our electricity consumption and associated environmental impacts a high priority in our fiber-optic network operations. The measures that need to be taken to achieve this are developed on the basis of the data from the EMS.

1&1 Versatel’s environmental management activities are centered on its major technical locations, which were identified in an ABC analysis as being the largest consumers of electricity. A substantial 46.55% (2019: 42.78%) of total electricity consumption was attributable to 1&1 Versatel’s 30 largest locations in the 2020 reporting period. The remaining 53.45% (2019: 57.22%) was distributed across roughly 2,384 (2019: 2,334) own and rented locations.(1) This is why the EMS is deployed at these major locations in order to hit optimization targets, and manage consumption and cost trends. The air conditioning technology in use at our major locations in Dortmund, Bielefeld, Frankfurt, and Essen was optimized in the reporting period. We installed direct and indirect free cooling systems here that utilize cool outside air where possible for air conditioning. We also installed more effective UPS systems at a number of locations and converted the lighting at many sites to, state-of-the-art, efficient LED versions.

(1) The figures for individual locations for which no readings are currently available are based on estimates.

In addition to the electricity that it sources itself at its 30 largest and other locations, 1&1 Versatel is connecting its fiber-optic network with fiber-optic networks and colocation sites belonging to well-known regional city carriers and network operators and, to an increasing extent, with Deutsche Telekom’s BNGs (Broadband Network Gateways). In these cases, 1&1 Versatel does not manage electricity purchasing itself.

The following table shows 1&1 Versatel’s electricity consumption and the locations where 1&1 Versatel is a user.

Electricity consumption and carbon emissions for the fiber-optic network(1)

2018(2)

2019

2020(2)

Own electricity consumption in MWh(3)

39,707.16

34,614.09

Electricity consumption at business partners caused by 1&1 Versatel(3) in MWh

19,892.02

22,051.51

Total electricity consumption in MWh

60,113.43

59,599.17

56,665.61

CO2 equivalents in tonnes (Scope 2; own electricity consumption)(4)

18,821.19

13,880.25

CO2 equivalents in tonnes (Scope 3; own electricity consumption)(4)

9,428.82

8,842.66

Total carbon emissions in metric tonnes

28,493.77

28,250.01

22,722.91

(1) The figures for fiscal year 2020 cover electricity consumption for all 1&1 Versatel’s technical locations, plus locations where 1&1 Versatel is a user. The electricity consumption figures for and 2019 also include electricity consumption at 1&1 Versatel's office locations, since this data was not available separately. Electricity consumption did not increase despite the expansion of the network, as it is being steadily reduced by efficiency measures.

(2) The figures for individual locations for which no readings are currently available are based on estimates; there is a delay before the data for energy consumption given for the years in question is allocated to specific time periods due to fluctuations in the reading and billing dates. Consequently, figures for prior periods may change after the report is published; as a result the figures for 2018 have been adjusted.

(3) The figures have only been available in this form since 2019. Although a proportion of the electricity consumed is green electricity, this cannot be quantified exactly; as a result, no figures for this have been given.

(4) Estimate based on the data provided by the Umweltbundesamt (the Federal Environmental Agency) in 2019 for CO2 emissions per kilowatt hour in the German electricity mix (2018, 2019: 474 grams; 2020: 401 grams). Actual emissions are lower due to the use of green electricity, although the proportion attributable to the latter cannot be quantified exactly.

Energy Consumption by Our Office Buildings

Relatively speaking, our office buildings account for a very small proportion of our energy consumption. In addition, a significant share of the electricity purchased in the reporting period was certified green electricity, something that reduces the potential environmental impact.

Energy consumption and carbon emissions for
office buildings

2018

2019

2020

Electricity consumption(1) in MWh

7,430.57

7,068.83

6,556.84

Heating consumption (natural gas and heating oil)(2) in MWh

2,886.65

2,859.01

3,130.86

Total energy consumption in MWh

10,317.22

9,927.84

9,687.70

CO2 equivalents from electricity consumption in tonnes (Scope 2)

1,367.90

451.30

418.86

CO2 equivalents from heating consumption in tonnes
(Scope 1)

593.39

587.52

640.61

Total carbon emissions in tonnes(3)

1,961.29

1,038.82

1,059.47

(1) The figures relate to the offices and related infrastructure at United Internet’s locations in Montabaur, Karlsruhe, Munich, and Vienna and to 1&1 Drillisch's locations in Maintal, Krefeld, Dresden, Nuremberg, Munich, and Münster, which are supplied directly by the utilities concerned. In addition, the United Internet locations in Düsseldorf and Hamburg are included as from 2019 onwards, and its Berlin location is included as from 2020. The figures for the 1&1 Drillisch companies also include electricity consumption by the 1&1 Drillisch data centers (this data has not been disclosed separately to date). Not all figures were available by the editorial deadline for this report. The figures for the 2020 reporting period include consumption at 1&1 Versatel’s offices for the first time. In previous years, this information was reported together with electricity consumption for the fiber-optic network (see page 84), since it was not possible to capture the data separately.

(2) The information for natural gas relates to the Montabaur site, including the logistics center and (as from 2020) Vienna, plus (for 2018) the 1&1 Drillisch locations in Maintal, Krefeld, and Münster; as from 2019, it relates to Maintal only. Heating oil relates to one location in Karlsruhe. Not all figures were available by the editorial deadline for this report.

(3) Calculated using the utilities’ data on CO2 emissions per kilowatt hour, the conversion factors provided by DEFRA (the UK’s Department for Environment, Food and Rural Affairs) for 2018, 2019, and 2020 and the data provided by the Umweltbundesamt (the Federal Environmental Agency) published in 2019 for CO2 emissions per kilowatt hour in the German electricity mix (2018, 2019: 474 grams; 2020: 401 grams). Green electricity, which does not result in any CO2 emissions, was used in some cases.