EN 50600 is a European standard for data centers that lays down comprehensive requirements for planning, construction, and operation. These cover structural design, power source, air conditioning, wiring, safety systems, availability tiers, and key performance indicators (KPIs). EN 50600 was put forward by the European organization CENELEC (European Committee for Electrotechnical Standardization) in 2011. The first sections were then published in 2012, followed by repeated additions and updates. These also included the fourth part of EN 50600, which defines the KPIs for data centers in terms of energy efficiency and emissions.
The EN 50600 KPI system makes it easier to compare sustainability factors and promotes the spread of efficient infrastructures. To improve comparability, when examining the KPIs of this fourth section we should also consider the levels of redundancy for the power source and distribution based on the various data center tiers, as set out in the second part of EN 50600. The individual KPIs cover all major aspects of data center operation. In 2018, the content of European standard EN 50600 was merged with the content of ISO 30134 to create the new international standard ISO/IEC 22237.
Below, we explain how the KPIs are measured and calculated and what target values are in the green zone. The starting point for evaluating the efficiency of a data center is energy consumption, which depends on the installed IT capacity on the one hand and on infrastructural components such as cooling and the UPS on the other hand.
1. Power Usage Effectiveness (PUE)
First of all, EN 50600-4-2 mentions the familiar Power Usage Effectiveness (PUE) metric. PUE was developed by The Green Grid as far back as 2007 and compares the data center’s total energy consumption with the energy consumption of the IT equipment. The ideal PUE would be 1.0, because in this case the total energy consumption would be used by the IT components. PUE is calculated as follows:
PUE = Total Facility Energy (kWh) / IT Equipment Energy (kWh)
There are several categories for measuring PUE, which correspond to different measurement points. The category number is added to the PUE as a subscript figure, e.g., PUE1 for a Category 1 measurement. Category 1 determines the energy consumption of the data center as a whole as read at the consumption meter and the IT equipment load at the UPS system output in kWh. Category 2 measures IT consumption at the PDU, while Category 3 records IT consumption directly at the server input and is therefore considered the most accurate method, because all impact of losses en route to the server is eliminated.
However, it isn’t enough to consider energy consumption alone. For sustainable data center operation, it is just as important to achieve an electricity mix with the highest possible share of renewable energy.
2. Renewable Energy Factor (REF)
The next KPI defined in EN 50600-4-3 is the Renewable Energy Factor (REF). This constitutes the proportion of the data center’s energy consumption from renewable sources and is calculated using the following formula:
REF = ERen (kWh) / EDC (kWh)
The share of renewable energy (ERen) may come from three sources:
- A portion of energy defined as renewable by the utility provider (REF * amount of energy drawn in kWh)
- Energy generated on-site (total amount in kWh, if this is used in the data center (DC) and does not exceed the data center’s consumption)
- Directly purchased Renewable Energy Certificates (acquired amount in kWh)
These amounts of energy are added together to form the overall value ERen. The ideal REF would be 1, which would mean that the data center’s entire energy needs are met by renewable energy.
To ensure compliance with forthcoming legislation, the energy used may not simply be expelled from the data center as waste heat; instead, a proportion must be repurposed (heating neighboring properties, feeding into heating networks).
3. Energy Reuse Factor (ERF)
The Energy Reuse Factor (ERF) set out in EN 50600-4-6 is a KPI describing the proportion of the energy used by the data center that is repurposed. When calculating the ERF, the repurposed energy may not be used directly by the data center, however, but in areas clearly beyond its boundaries – for example in an office building that is adjacent to the data center and not included in its metered energy consumption. The ERF is calculated as follows:
ERF = EReuse (kWh) / EIT (kWh)
EIT is the energy consumption measured at the PDUs that supply electricity to the IT components.
EReuse is measured at the energy handover point that is closest to the data center source. When waste heat is recovered by heat exchangers, the amount of heat used must be the same as that transferred to the waste heat recovery network.
Despite the incorporation of waste heat recovery, cooling remains the largest consumer after the IT equipment. To ensure sustainable operation, here we need systems that are highly efficient and can therefore remove the greatest possible amount of heat from the energy used.
4. Cooling Efficiency Ratio (CER)
To assess the efficiency of cooling systems, EN 50600-4-7 defines the Cooling Efficiency Ratio (CER) KPI, which replaces the annual performance factor used previously. This indicates the amount of heat removed relative to the required cooling capacity. The CER is calculated as follows:
CER = removed heat (kWh) / energy consumption of cooling equipment (kWh)
Data center certification with the Blue Angel, for example, helps to clarify what values are realistic here. A data center built today would have to achieve a CER of more than 8. For an older data center that went into operation on or before 12/31/2014, the stipulated figure is more than 5.
Apart from the energy efficiency of IT and cooling equipment, the lowest possible CO2 emissions are also a decisive element in achieving sustainable data center operation.
5. Carbon Usage Effectiveness (CUE)
Carbon Usage Effectiveness (CUE) is a KPI defined in EN 50600-4-8 to assess how effectively CO2 emissions are avoided. It measures CO2 emissions relative to the energy consumption of the IT equipment. However, CUE only registers direct Scope 1 emissions. CUE is calculated as follows:
CUE (kgCO2eq/kWh) = CO2 emissions (kgCO2eq) / energy consumption of IT equipment (kWh)
To calculate CUE, data center emissions are divided by the energy consumption of the IT equipment, as is the case with Power Usage Effectiveness (PUE). An ideal CUE would be 0.
In a growing number of regions, data center water usage is also an increasingly hot topic in times of water scarcity, and it is essential to take account of this factor in data center planning and operation.
6. Water Usage Effectiveness (WUE)
Water Usage Effectiveness (WUE) is a KPI set out in EN 50600-4-9 that defines the ratio between the data center’s total annual water usage and the energy consumption of the IT equipment. An ideal WUE would be 0.0 which would, however, require a data center to dispense entirely with the use of fresh water for cooling. WUE is calculated as follows:
WUE (m³/kWh) = total water usage (m³) / IT energy consumption (kWh)
How is WUE measured?
The data center’s total water usage is measured directly at the supplying water meter as water is drawn from the network. If water is taken from a lake or river, the amount used must be measured based on the flow into the cooling systems or adiabatic modules. The energy consumption of the IT equipment is already known from the PUE calculation.
The factors considered above are also directly influenced by the level of redundancy available in the data center. Therefore, any data centers compared should ideally always be in the same tier.
7. Power source and distribution redundancy according to EN 50600-2-2
If we wish to improve our comparison of data center efficiency above and beyond the above KPIs, the levels of redundancy for power source and distribution specified in EN 50600-2-2 for the different tiers should also be included in our calculations. Ultimately, it would make little sense to compare the power source and distribution of a Tier 1 data center with the lowest uptime availability with a Tier 4 data center with the maximum availability, because a considerably higher level of redundancy is required for a Tier 4 data center, and this naturally also impacts on energy efficiency. The tiers are defined as follows:
- Tier 1: Low availability and no redundancy, shutdown is necessary in the event of maintenance and malfunctions
- Tier 2: Medium availability, important components have redundancy and can continue to be operated during maintenance; other components must be shut down in the event of malfunctions and maintenance
- Tier 3: High availability, multiple paths and redundancy for important components; routine maintenance can be performed during ongoing operation
- Tier 4: Very high availability, multiple paths and redundancy for all components; routine maintenance can be performed during ongoing operation; the data center is fault-tolerant during normal operation
The differences can be simply explained by comparing Tier 2 (N+1 redundancy) and Tier 4 (2N redundancy). If the Tier 2 data center has four UPS systems to back up the available load, just one more system would have to be added to achieve N+1 redundancy. If the Tier 4 data center also needs four UPS systems to cover the required load, four further systems would have to be added to achieve 2N redundancy, which naturally consume far more energy than one. As this effect is replicated with many other components, there are dramatic differences in the consumption of IT and infrastructural components, even though the IT load is the same.
Conclusion
The KPI system presented in EN 50600 provides data center operators with important tools for comparing the energy efficiency of data centers in Europe. Furthermore, EN 50600 already helps data center operators ensure the correct measurement of all KPIs now, before the introduction of legal requirements such as the EU CSRD (Corporate Sustainability Reporting Directive) and the German Energy Ordinance (EnEV). Numerous organizations (e.g. TÜV-IT) also offer EN 50600 or ISO/IEC 22237 certification of data centers and provide customers and partners with peace of mind that a data center has been planned in accordance with generally recognized basic principles and that its operation takes full account of both reliability and sustainability aspects.