Global greenhouse gas emissions from the tech sector are on par or larger than the aviation industry, at around 3% for ICT and 2% for aviation respectively. Within ICT, data centers consume around 1% of greenhouse gas emissions and global electricity usage. Currently, most cloud providers do not disclose energy or carbon emissions from cloud usage to their customers (at an aggregate or individual level), which can be a challenge for organizations who want to baseline and reduce their carbon footprint. This application is a starting point to enable organizations to have greater visibility into their emissions across multiple cloud providers.

There are currently no guidelines for reporting Scope 3 emissions as part of the Greenhouse Gas (GHG) Protocol, which cloud provider usage would fall under. However we hope more and more organizations report on both location-based and market-based emissions from cloud usage. To that end, this application’s focus is on providing a location-based cloud emissions estimate, and we welcome contributions that could aid with market-based reporting to include energy attributes such as RECs or power purchasing agreements.

This application pulls usage data (compute, storage, networking, etc) from major cloud providers and calculates estimated energy (Watt-Hours) and greenhouse gas emissions expressed as carbon dioxide equivalents (metric tons CO2e). We display these visualizations in a dashboard for developers, sustainability leaders and other stakeholders in an organization to view and take action. It currently supports AWS, Google Cloud and Microsoft Azure.

We calculate our CO2e estimates with this formula:

(Cloud provider service usage) x (Cloud energy conversion factors [kWh]) x (Cloud provider Power Usage Effectiveness (PUE)) x (grid emissions factors [metric tons CO2e])

Our approach builds upon Etsy's Cloud Jewels (cloud energy conversion factors) for estimating CO2e emissions for cloud compute and storage services, with the addition of networking and memory usage. We similarly use point estimates without confidence intervals due to the experimental nature of the project, which are not meant as a replacement for data from cloud providers and we cannot guarantee their accuracy.

We encourage and welcome any improvements and extensions to both the methodology and software.We aim to keep this approach up to date, including using the most recent publicly available data.

Longer Version#

A note on our approach#

Our application is designed to be a starting point which can be extended and customized for individual organizations' needs. Every organization will have a different cloud setup and tech stack, so we are using domain driven design to separate the estimation logic from both the data input source (e.g. cloud APIs, on-premise or co-located data centers) and the output source (e.g front-end dashboard, CSV, etc) so new inputs and outputs can easily be added.

Options for cloud usage and cost data source#

We support two approaches to gathering usage and cost data for different cloud providers. One approach gives a more holistic understanding of your emissions whereas the other prioritizes accuracy:

1. Using Billing Data for Cloud Usage (Holistic)#

By default, we query cloud provider billing and usage APIs:

  • AWS Cost and Usage Reports with Amazon Athena
  • GCP Billing Export Table using BigQuery.
  • Azure Consumption Management API

This pulls usage and cost data from all linked accounts in your AWS, GCP, or Azure Organization. This approach provides us with a more holistic estimation of your cloud energy and carbon consumption, but may be less accurate as we use an average constant (rather than measured) CPU Utilization.

Before estimating the energy and carbon emission, we validate whether a given usage is Compute, Storage, Networking, Memory or Unknown, with Unknown being unused in the estimation formula. You can see our classifications of these usage types in following files:

  • AWS: packages/core/src/services/aws/CostAndUsageTypes.ts
  • GCP: packages/core/src/services/gcp/BillingExportTypes.ts
  • Azure: packages/core/src/services/azure/ConsumptionTypes.ts

The process by which we classified the usage types is:

  1. Consider the pricing (AWS) or usage (GCP, Azure) unit: if it is hours or seconds, it is likely to be a Compute usage type. If it is byte-seconds or GigaByte-Months, it is likely to be Storage, and if it is bytes of Gigabytes it is likely to be networking. Most other units are ignored.
  2. We then further validate whether a line item is Compute, Storage, or Networking by looking at the more detailed usage type. E.g. if it contains content like “RAM”, it would be ignored.

You can see more details about this logic in following files:

  • AWS: packages/core/src/services/aws/CostAndUsageReports.ts
  • GCP: packages/core/src/services/gcp/BillingExportTable.ts
  • Azure: packages/core/src/services/azure/ConsumptionManagement.ts

We welcome additions, improvements or suggested changes to these classifications or the process.

The way we determine total vCPU Hours for the compute estimation varies for each cloud provider. For AWS and Azure we multiply the usage amount by the product vCPUs, because our understanding is that the usage amount doesn’t include the vCPU count for a given usage row. For GCP, our understanding is that the vCPU count is included in the usage amount for a given row, so we simply use the usage amount by itself.

For AWS Savings Plans, we only include the line item type SavingsPlanCoveredUsage because our understanding is that the other Savings Plans line item types refer to fees or discounts in the form of refunds.

When calculating total kilowatt hours for AWS Lambda service using Billing Data (Holistic), we are assuming that MemorySetInMB will be 1792, and since we will then divide this by the constant 1792, we just don't include it in the calculation.

For details on some of the specific usage type classifications, please view the Classifying Usage Types page.

2. Using Cloud Usage APIs for Cloud Usage (Higher Accuracy)#

This approach utilizes the AWS CloudWatch and Cost Explore APIs, and the GCP Cloud Monitoring API to pull usage and cost data. We achieve this by looping through the accounts and then making the API calls on each account for the regions and services set in the application configuration. We retrieve an hourly granularity for usage and daily granularity for cost. This approach is arguably more accurate as we use the actual CPU utilization percentage in the emission estimation for compute usage, but is confined to the services that have been implemented so far in the application. We currently only support compute and storage for this approach.

The cloud providers and services currently supported with this approach are:


  • EC2 (compute)
  • Lambda (compute)
  • EBS (storage)
  • RDS (compute & storage)
  • S3 (storage)
  • Elasticache (compute)


  • Compute Engine (compute)

Azure is not currently supported for this approach.

Energy Estimate (Watt-Hours)#

In order to estimate energy used by cloud providers we are leveraging the methodology that Etsy created called “Cloud Jewels” to determine energy coefficients (kWh) for cloud service compute and storage usage. In addition, we’ve added energy estimation for networking and memory usage. The application also doesn’t currently include estimations for cloud GPU usage, but this is on the roadmap. You can see a summary of all our energy coefficients in Appendix I below.

We look at the servers used by cloud providers on their website and reference their energy usage from both the SPECPower database and the 2016 US Data Center Energy Usage Report. Etsy looked into GCP servers and we have additionally looked into AWS and Azure servers; see list of processors below in Appendix II. Of course this does not account for any custom processors (such as AWS has) however this is the best information we found publicly available.


For Compute estimation, we follow the same formula as Cloud Jewels, which can be broken down into 2 steps. (We recommend reading their article for a deeper explanation).

First, we calculate Average Watts which is the average compute energy at a moment in time. When a server is idle, it still takes some power to run it (Minimum Watts). As the server utilization increases the amount of power consumed increases too. The total energy used is the min watts plus the watts from additional server usage (average per hour).

Average Watts = Min Watts + Avg vCPU Utilization * (Max Watts - Min Watts)

Second, we then translate this into total Watt Hours based on the amount of time servers are being used, or virtual CPU hours.

Compute Watt-Hours = Average Watts * vCPU Hours

Here are the input data sources for the variables in the formula, and context on where we have sourced them:

  • Min Watts (constant) - This is dependent on the CPU processor used by the Cloud provider to host the virtual machines. Based on publicly available information about which CPUs cloud providers use, we looked up the SPECPower database to determine this constant per processor micro-architecture.
  • Max Watts (constant) - Same as Min Watts, above.
  • Avg vCPU Utilization (variable or constant) - This is either pulled from the cloud provider APIs (see above) or falls back to 50% (see below).
  • vCPU Hours (variable) - This is pulled from the cloud usage APIs or billing data .

When the actual Avg vCPU Utilization for a given time period isn’t available from a cloud provider's API, we fall back to a projected estimate for the average server utilization of servers in hyperscale data centers in 2020 of 50%, from the 2016 U.S. Data Center Energy Usage Report. For example, this may occur for AWS EC2 instances that are terminated over 2 weeks ago from when the application first queries an AWS Account.

When we know what underlying processor micro-architecture or group of micro-architectures are used for a given cloud provider virtual machine, we use the min and max watts for that micro-architecture or the average of a group of micro-architectures. When a group of micro-architectures includes either Ivy Bridge or Sandy Bridge, we use the median of that group. This is because we treat those micro-architectures as outliers due to their comparatively high min/max watts in order to not overestimate. See Appendix II for this list of processors and micro-architectures with the min and max watts.

When we don’t know the underlying processor micro-architecture, we use the average of all micro-architectures used by that cloud provider. Here are those averages per cloud provider:


  • Min Watts: 0.71
  • Max Matts: 3.46


  • Min Watts: 1.34
  • Max Watts: 4.98


  • Min Watts: 0.77
  • Max Matts: 3.74

A note on AWS Lambda Compute Estimates#

In the case of AWS Lambda, AWS does not provide metrics for CPU Utilization and number of vCPU hours, so we need to take an alternative approach.

Lambdas can consume between 128MB and 10,240MB in 1MB increments [source]. At 1,792MB a lambda has an equivalent of one vCPU [source]. Above this another vCPU is assigned up to a total of 6 vCPUs, and we can estimate the number of vCPUs allocated to a lambda function as a ratio of the allocated memory over 1,792MB.

Given this, the formula we derive is:

Total Watt-Hours = Average Watts X Running Time (Hours) X Estimated number of vCPUs


Average Watts = Min Watts + 50% (Average for hyperscale data centers) * (Max Watts - Min Watts)

Running Time = Lambda execution time / 3600 seconds

Estimated number of vCPUs = Lambda memory allocated (MB) / 1,792 MB

The execution time and memory allocated are both pulled from the Cost and Usage Reports or CloudWatch Lambda Logs.

A note on AWS Aurora Serverless Compute Estimates#

In the case of AWS Aurora Serverless using the Cost and Usage Reports, the pricing unit is ACU-Hrs. 1 ACU has approximately 2 GB of memory with corresponding CPU and networking, similar to what is used in Aurora user-provisioned instances [source]. Looking at the most recent Aurora instance classes, the number of vCPUs provisioned is directly proportional to the amount of memory - roughly one eighth [source].

Given that, for every ACU Hour, we estimate that 0.25 vCPUs are provisioned (2 GB of memory divided by 8). So for Compute estimations, the application treats every 4 ACU-Hours as one vCPU Hour.


For storage, we also follow the same methodology as Cloud Jewels by deriving a Wh/Tbh coefficient for both HDD and SSD storage types. However we updated the coefficients to be more accurate for the year 2020, based projections from the same 2016 U.S Data Center Usage Report.

Here is the estimated HDD energy usage:

  • HDD average capacity in 2020 = 10 Terabytes per disk
  • Average wattage per disk for 2020 = 6.5 Watts per disk

Watts per Terabyte = Watts per disk / Terabytes per disk: 6.65 W / 10 TB = 0.65 Watt-Hours per Terabyte-Hour for HDD

Here is the estimated SSD energy usage:

  • SSD average capacity in 2020 = 5 Terabytes per disk
  • Average wattage per disk for 2020 = 6 Watts per disk

Watts per terabyte = Watts per disk / Terabytes per disk: 6 W / 5 TB = 1.2 Watt-Hours per Terabyte-Hour for SSD

When it comes to measuring the Terabytes from cloud providers, we query for the allocated bytes rather than the utilized bytes, because this is a more accurate reflection of the energy needed to support that usage. For example, an organization may have a 20 Gigabyte AWS EBS Volume allocated, but is only utilizing 2 Gigabytes for this block storage device. In this case we would use 20 GBs in the energy estimation formula for EBS storage.

Storage Services Replication Factors#

In order to achieve adequate durability and availability for data stores and to ensure better redundancy in the case of service outages, most cloud provider storage and database services automatically replicate your data as well as any associated compute and memory resources. Sometimes this is within an individual data center, other times it is within a geographical location or across multiple geographical locations.

Because of this, the actual infrastructure (and energy) required to provide a given amount of storage can be multiple times more than what might be provided by cloud providers as the allocated amount of usage. After analyzing cloud storage and database services, we have determined a number of “replication factors” to take this into account, which you can see the details of in this spreadsheet. These replication factors are applied to the total energy and CO2e estimate for each storage or database service (inclusive of all types of resources replicated).



Currently, our application takes into account only the data exchanged between different geographical data centers.

For networking, it is safe to assume that the electricity used to power the internal network is close to 0, or at least negligible compared to the electricity required to power servers. We also have chosen to ignore traffic that leaves a data center to provide end-users with services, because this traffic is usually handled by CDN providers (Content delivery network) which are not necessarily the same provider as for cloud computing. This traffic is also dependent on the behavior of end-users. That said, we would welcome contributions to be able to include this in our approach.

Studies to date#

There have not been many studies that deal specifically with estimating the electricity impact of exchanging data across data-centers. Most studies focus on estimating the impact of end-user traffic from the data center to the mobile phone; integrating the scope of the core network (what we are interested in), the local access to internet (optical fiber, copper, or 3G/4G/5G) and eventually the connection to the phone (WiFi or 4G).

On top of that, these studies use different methodologies and end up with results with orders of magnitude in differences. See appendix III below for a summary of the most recent studies. Note that it is very hard to find recent studies that provide an estimation for optical fiber networks, the scope we are interested in.

Chosen coefficient#

It is safe to assume hyper-scale cloud providers have a very energy efficient network between their data centers with their own optical fiber networks and submarine cable source. Data exchanges between data-centers are also done with a very high bitrate (~100 GbE -> 100 Gbps), thus being the most efficient use-case. Given these assumptions, we have decided to use the smallest coefficient available to date: 0.001 kWh/Gb. Again, we welcome feedback or contributions to improve this coefficient.

We want to thank @martin-laurent for providing this research and recommended coefficient.


Chosen Coefficient#

For the purpose of estimating energy consumption for memory, we are assuming hyper-scale cloud providers are utilizing the more efficient memory systems on the market: DDR4 or potentially DDR5. Two memory manufacturers have provided some information about the energy profile of DDR4 memory systems. Crucial says that “As a rule of thumb, however, you want to allocate around 3 watts of power for every 8GB of DDR3 or DDR4 memory,” which amounts to ~0.375 W/GB. Micron provides a power model that states “... each DRAM will consume approximately 408.3mW of total power,” which equals a power consumption of ~0.4083 W/GB. Given this information, we have decided to take the average of both these figures, and go with 0.392 W/GB, or 0.000392 kWh/GB. We want to acknowledge that this is a complex subject with limited available data, and welcome additional research or studies to improve this coefficient.

Applying the coefficient#

When we utilize the SPECPower Database for estimating compute (above), this also includes some level of energy estimation for memory, because the rows in that database represent the min and max watts for entire servers, which includes memory usage. Because of this, we only want to apply an additional estimation for memory when the instance you have selected from a given cloud provider has a higher amount of memory allocated per physical CPU than what is provided in the SPECPower database.

The way in which we apply this coefficient is slightly different per cloud provider:


Our approach to estimating memory for AWS is as follows:

  1. For each microarchitecture in the SPEC Power Database we have determined the average memory (GB) / physical CPU (Chips)
  2. For each AWS instance family, we have determined how many physical CPUs we believe are provisioned for the instance type that is roughly equivalent to an entire server, like the rows in the SPEC Power Database. This is often the largest instance in a family, e.g. a “metal” instance.
  3. For each AWS instance family, we calculate the difference between the GB / physical CPU from the associated microarchitectures in the SPECPower Database, and the GB / physical CPUs for the comparable instance size, referred to in step 2.
  4. If this difference (in GB) is greater than zero, we know an additional memory estimation should be added for this usage row. The amount of kilo-watt hours added for memory is directly proportional to the size of the instance - this is because memory allocated scales linearly within a family. We multiply this proportional amount of additional memory by the chosen coefficient to get kilo-watts hours.
  5. If the GB / physical CPU difference is less than or equal to zero, we ignore the usage. We also ignore the usage for any burstable instances, because there is no instance size comparable (large enough) to the SPECPower Database rows.
  6. You can see a full list of these instance types and calculations for memory here.

Here is that formula summarized:

Kilowatt hours = Memory (GB) exceeding SPECPower database average x Memory coefficient x usage amount (Hours)


Because Google Cloud provides memory specific usage rows in the Billing Export data, the approach is a lot simpler. We take the usage amount in byte-seconds, convert this to gigabyte-hours, then multiply it by the chosen memory coefficient to get the kilowatt hours.

Here is that formula summarized:

Kilowatt hours = Memory usage (GB-Hours) x Memory coefficient

This approach may be slightly overestimating the total energy and carbon emissions for GCP, because of the note above about the min/max watts coefficients from the SPECPower database including some level of memory usage for compute estimates. We think this approach is still more accurate than not taking into account the memory usage rows at all, and are researching a better way to factor this in for compute estimates. We welcome any and all feedback/suggestions on how to improve this.


Our approach to estimating memory for Azure is very similar to AWS. When it comes to compute instances, it is exactly the same approach, and here is a spreadsheet with the full list of instance types and calculations for memory.

In addition to compute instance usage rows, we have found that Azure also has some usage rows that correspond directly to memory usage, e.g. “Memory Duration”. In those cases, we apply the same simplified formula as GCP (above).

We want to thank @benjamindavyteads for collaborating with us on this approach, and providing much of the research behind it, including the count of physical CPUs for AWS instances that are roughly equivalent to the SPEC Power database rows.

Power Usage Effectiveness#

After estimating the kilowatt hours for compute, storage and networking, we need to multiply this by the cloud provider Power Usage Effectiveness (PUE). PUE is a score of how energy efficient a data center is, with the lowest possible score of 1 meaning all energy consumed goes directly to powering the servers and none is being wasted on cooling. This is based on publicly available information provided by the cloud providers. In the case of GCP, they publish their PUE. In the case of AWS, we have made a conservative guess based on public information. Microsoft's Sustainability team have provided a statement1 as to the PUE for the Azure datacenters.

Here are the cloud provider PUEs being used:

  • AWS: 1.135
  • GCP: 1.1
  • Azure: 1.185

[1] "Organizations are increasingly using the power usage effectiveness (PUE) metric to evaluate the energy efficiency of potential cloud service providers. PUE is essentially the energy going into the datacenter divided by the energy used by the servers. Differing methodologies across the industry for obtaining this metric makes precise comparison across providers difficult. Nevertheless, we have achieved significant improvements as we have continued to evolve both our designs and our operations to achieve among the best efficiency in the industry. Currently, our weighted owned and operated fleet-wide PUE trailing 12-month average is 1.185, and our latest designs are achieving an annual PUE of 1.125."

Carbon Estimates (CO2e)#

Once we have the estimated kilowatt hours for usage of a given cloud provider, we then convert that into estimated CO2e using publicly available data on emission factors for a given electricity grid based on the mix of local energy sources. We do this based on the cloud provider datacenter region that each service is running in.


Google has published the grid carbon intensity for their GCP regions, so in this case we use their published factors.

AWS & Azure:

In the United States, we use the EPA’s eGRID2019 Data that provides NERC region specific emission factors annual for CO2e. We decided to use the NERC region emission factors rather than the more granular eGRID subregion or state emissions factors because we feel that it better represents the energy consumed by data centers, rather than the energy produced in a given state/subregion which those metrics would more adequately reflect. Outside the US, we generally use’s country specific grid emissions factors report. For most of Europe, however, we use EEA emissions factors. In the case of Singapore, we get the data from the Energy Market Authority’s electricity grid emissions factors, and for Taiwan we got it from energypedia as neither are included in the report.

You can see the full list of emissions factors in Appendix IV below.

We understand this is a rough estimated conversion as these are only averages over a given year that is pre-2020, and they also don’t take into account time of day. We welcome improvements to this, for example electrictyMap API provides hourly historical and forecasted electricity emissions data for a fee.

Appendix I: Energy Coefficients:#


  • Average Minimum Watts (0% Cpu Utilization): 0.71
  • Average Maximum Watts (100% Cpu Utilization): 3.46
  • Average CPU Utilization for hyperscale data centers: 50%
  • HDD Storage Watt Hours / Terabyte: 0.65
  • SSD Storage Watt Hours / Terabyte: 1.2
  • Networking Kilowatt Hours / Gigabyte: 0.001
  • Memory Kilowatt Hours / Gigabyte: 0.000392
  • Average PUE: 1.135


  • Average Minimum Watts (0% Cpu Utilization): 1.34
  • Average Maximum Watts (100% Cpu Utilization): 4.98
  • Average CPU Utilization for hyperscale data centers: 50%
  • HDD Storage Watt Hours / Terabyte: 0.65
  • SSD Storage Watt Hours / Terabyte: 1.2
  • Networking Kilowatt Hours / Gigabyte: 0.001
  • Memory Kilowatt Hours / Gigabyte: 0.000392
  • Average PUE: 1.1


  • Average Minimum Watts (0% Cpu Utilization): 0.77
  • Average Maximum Watts (100% Cpu Utilization): 3.74
  • Average CPU Utilization for hyperscale data centers: 50%
  • HDD Storage Watt Hours / Terabyte: 0.65
  • SSD Storage Watt Hours / Terabyte: 1.2
  • Networking Kilowatt Hours / Gigabyte: 0.001
  • Memory Kilowatt Hours / Gigabyte: 0.000392
  • Average PUE: 1.125

Appendix II: Cloud provider compute processors and micro-architectures:#

You can see the full list of AWS, GCP and Azure microarchitectures and min/max coefficients used for Compute in this spreadsheet.

Note on AWS Graviton 2 Processor:

When it comes to the AWS Graviton 2 custom processor, it is likely more efficient than other processors however we are yet to find any reliable min or max watts values. For the time being, we apply the lowest min/max watts for any microarchitecture: AMD EPYC 2nd Gen. The same is true for the GB / physical chip used to estimate energy for memory usage.

Appendix III: Recent Networking studies#

StudyScopeYear (data applied)Energy intensity kWh/GB
Methodology To Model the Energy and Greenhouse Gas Emissions of Electronic Software Distributions Williams & Tangend-to-end20100.13
Understanding the environmental costs of fixed line networking Krugend-to-end20120.14
Understanding the environmental costs of fixed line networking KrugInternet core20120.04
Electricity Intensity of Internet Data Transmission: Untangling the Estimates Aslan et alend-to-end20150.06
Electricity Intensity of Internet Data Transmission: Untangling the Estimates Aslan et alInternet core20150.004
New perspectives on internet electricity use in 2030, A. AndraeFixed access network20200.07 - 0.055
Talk by J. MalmodinFixed broadband network?0.1 - 0.001 (depending on the bitrate)

Appendix IV: Grid emissions factors:#


RegionCountryNERC RegionCO2e (metric ton/kWh)Source
us-east-1United StatesSERC0.000415755EPA
us-east-2United StatesRFC0.000440187EPA
us-west-1United StatesWECC0.000350861EPA
us-west-2United StatesWECC0.000350861EPA
us-gov-east-1United StatesSERC0.000415755EPA
us-gov-west-1United StatesWECC0.000350861EPA
ap-southeast-1Singapore0.0004085EMA Singapore


RegionLocationCO2e (metric ton/kWh)Source
us-east1South Carolina0.0005Google
us-east4Northern Virginia0.000383Google
us-west2Los Angeles0.000248Google
us-west3Salt Lake City0.000561Google
us-west4Las Vegas0.000491Google
asia-east2Hong Kong0.000626Google
asia-northeast3South Korea0.00054Google


Note: The application currently only supports a subset of Azure regions that are used by ThoughtWorks. This is because the syntax in which they are returned from the Azure Consumption API doesn't always match what is listed in the Azure website. For example, the website says "Central US", but the API provides the region as "US Central". In the case of "UK South", it is the same on both the website and the API. For any Azure customers using other regions, we would love to know what syntax is returned by the API for your region(s) so that we can add support for them. You can email [email protected], or submit an issue or pull request.

RegionLocationNERC RegionCO2e (metric ton/kWh)Source
Central USIowaMRO0.00047223EPA
East USVirginiaSERC0.000415755EPA
East US 2VirginiaSERC0.000415755EPA
East US 3GeorgiaSERC0.000415755EPA
North Central USIllinoisRFC0.000440187EPA
South Central USTexasTRE0.000396293EPA
West Central USWyomingWECC0.000350861EPA
West USCaliforniaWECC0.000350861EPA
West US 2WashingtonWECC0.000350861EPA
West US 3ArizonaWECC0.000350861EPA
East AsiaHong
Southeast AsiaSingapore0.0004085EMA Singapore
North EuropeIreland0.000316EEA
West EuropeNetherlands0.00039EEA
UK SouthLondon0.000228EEA
UK WestCardiff0.000228EEA

Cloud Carbon Footprint is an open-source project, sponsored by ThoughtWorks Inc. under the Apache License, Version 2.0