Many businesses rent co-location (co-lo) space for their IT infrastructure, including servers, networking equipment, and storage. This approach allows companies to avoid the high costs and complexities of building and maintaining their own data centers while benefiting from professionally managed facilities.

‍

However, a key challenge arises when businesses need to report on their clean energy usage and carbon emissions. Since they do not own these facilities, they often lack direct metering data on the electricity consumption related to their rented space in that facility. Additionally, co-lo data centers may be reluctant to provide detailed usage information as this can impact their leverage in commercial negotiations with their businesses, potentially creating reporting challenges and compliance risks.

‍

With evolving reporting standards, such as the upcoming Corporate Sustainability Reporting Directive (CSRD)—which specifically addresses data centers in Scope 3 emissions reporting—companies must find accurate ways to gather energy usage data from these facilities.

‍

At Flexidao, we specialize in solving the data-gathering challenges, also those associated with co-lo spaces. We've worked with numerous businesses to develop practical solutions, avoiding the costly and resource draining process of installing and maintaining proprietary telemetry systems in each leased facility. 

‍

Below, we outline four possible methods to calculate electricity consumption for emissions reporting. Each provides a different level of accuracy and requires a different level of collaboration with the data center, which should be considered when deciding which one to implement. The graph below provides a useful overview.

‍

‍

‍

Option 1: Using Meter Readings and Power Usage Effectiveness (PUE)

‍

The most accurate method for calculating energy consumption involves using real meter readings combined with the realtime Power Usage Effectiveness (PUE) ratio of the data center. PUE is a key metric that quantifies data center efficiency by comparing total facility energy consumption to the energy used solely by IT equipment. A lower PUE indicates a more efficient facility, as a greater proportion of energy is dedicated to computing rather than overhead.

‍

If a co-lo data center provides meter readings for the electricity consumption associated with a leased space, businesses can calculate their IT energy use by dividing the total facility energy by the PUE ratio. This allows for precise emissions tracking and classification: electricity used for IT operations counts toward the reporting businesses Scope 2 emissions, while energy consumed by cooling, lighting, and other overheads of the data center falls under the businessesScope 3 emissions.

‍

Example Calculation:

• Total electricity usage: 240 kWh/day
• PUE: 1.4 (1 kWh for IT equipment, 0.4 kWh for overhead)
• IT Equipment Energy consumption: 171.43 kWh/day (Consumption to consider in Scope 2 or 3 emissions, depending on ownership of the hardware).
• Overhead Energy: 68.57 kWh/day (Consumption to consider in Scope 3 emissions)

While this method provides the most accurate results, it requires cooperation from the co-lo provider, which must share metering data and update PUE values regularly. Managing this information at scale demands robust data aggregation and processing, which is why many businesses partner with platforms like Flexidao to automate and simplify this process.

‍

Option 2: Estimating Consumption Based on Hardware Capacity

If a business knows the total hardware capacity it has in the co-lo, it can estimate electricity consumption based on the usage patterns of that equipment. This involves calculating the total consumption of servers, storage, and networking devices based on their specifications and workload patterns. Businesses can work with a third party like Flexidao, which has worked with partners before to convert hardware capacity into energy consumption estimates, ensuring calculations are as accurate as possible.

This approach requires no collaboration with the data center as the business should already have details of its hardware capacity, making it an independent and viable solution. It is also relatively accurate, as it is based on concrete capacity figures. 

However, a key limitation is that businesses may not always have a detailed and up-to-date inventory of the hardware they have deployed in the co-lo facility. Additionally, fluctuations in hardware usage can make static calculations less reflective of real-world consumption patterns.

‍

Option 3: Estimating Consumption Based on Leased Space

For companies that do not have precise data on the capacity of their IT hardware, an alternative approach is to estimate energy consumption based on the amount of space leased in the data center. Since data centers typically have standardized power allocations per square foot or per rack, businesses can use these allocations as a basis for calculating their energy use.

This method provides a general indication of electricity consumption but can result in overestimations or underestimations. For instance, not all rented space may be fully occupied by active IT equipment, leading to inflated energy consumption figures. To give an analogy, it is like estimating travel time based on speed limits rather than the actual driving speed of your car—it can provide a useful benchmark but will not give such an accurate insight.

On the plus side, this method requires minimal collaboration with the data center, as businesses already have access to their leasing agreements. Partnering with Flexidao can help refine these calculations by incorporating industry benchmarks and operational insights, improving accuracy without direct metering data.

‍

Option 4: Estimating Based on Total Facility Energy and Leased Percentage

In cases where businesses cannot access data on their hardware, leased space, or sub-metered electricity consumption, a last-resort approach is to estimate energy usage based on the total facility energy consumption and the percentage of the facility they occupy. By obtaining the total energy usage of the co-lo and determining what fraction of the space they are leasing, businesses can approximate their energy consumption.

However, this method is the least accurate of the four options. It assumes a uniform energy distribution across the facility, which may not reflect reality, as power usage can vary significantly between different areas of a data center. Additionally, data centers may be hesitant to disclose their total energy consumption, as this information could influence lease negotiations or reveal operational inefficiencies.

Despite its limitations, this method serves as a fallback option when no other data is available. Companies should explore alternative methods whenever possible to ensure more precise reporting.

‍

A Case-by-Case Approach

For businesses with multiple co-lo providers worldwide, a one-size-fits-all approach may not be feasible. Instead, a combination of the methods above may be necessary based on the data available at each site. Partnering with an expert like Flexidao can help companies navigate these challenges, ensuring they gather accurate energy usage data efficiently while complying with evolving sustainability regulations.

‍

Would you like to explore how we can tailor a solution for your business? Contact Flexidao to learn more about how we can help optimize your energy reporting strategy.