Simplifying the path to net-zero facilities with Siemens Decarbonization Business Optimizer powered by AWS

Siemens Financial Services (SFS) recognizes the complexity of decarbonizing buildings and financing the net-zero transition. This can be overwhelming for many small and medium-sized enterprises (SMEs). Sustainability transformations in SMEs is key for the United States to meet its committed emissions targets under The Paris Agreement, an international treaty that aims to reduce greenhouse gas emissions and limit global warming.

To help SMEs, SFS launched the Decarbonization Business Optimizer (DBO™), a free digital tool built in the Amazon Web Services (AWS) cloud, that empowers businesses of any size to embark on their decarbonization journey. SFS’s objective was to design an architecture that would be not only cost and energy efficient, but flexible and resilient to peaks in usage. Siemens Financial Services and Siemens Technology collaborated with AWS to optimize the Decarbonization Business Optimizer (DBO™) solution to improve scalability, optimize costs, and reduce the application’s carbon footprint. By leveraging serverless offerings by AWS, the team improved DBO so it can scale up or down based on user demand, without the need to provision or manage any of the underlying infrastructure.

In this blog, we will highlight what the DBO solution is and why it is important, the key design decisions made to improve the architecture on AWS, and the benefits that the serverless approach brought to the tool.

What is the Decarbonization Business Optimizer?

The Decarbonization Business Optimizer (DBO™) is a free digital tool that empowers businesses of any size to embark on their decarbonization journey. It is a comprehensive solution first providing a baseline carbon assessment, based on open, trusted data coming from the Environmental Protection Agency (EPA), the United States Department of Energy (US DOE), and the US Energy Information Administration (EIA). DBO™ then leverages advanced algorithms, including those from the National Renewable Energy Laboratory (NREL), to provide customized, cost-optimal decarbonization strategies, uncovering high-value pathways to achieve net-zero emissions. By removing initial knowledge barriers, the user-friendly tool enables businesses to analyze the potential benefits of these strategies, reduce costs, and maintain a competitive edge by facilitating the initial carbon assessment required by larger companies. DBO equips SMEs with data-driven insights and recommendations for implementing sustainable practices, paving the way for a greener, more resilient future. On September 24, 2024, during NYC Climate Week, the DBO tool made its public debut allowing anyone in the US to access and use it.

DBO builds an initial baseline carbon footprint assessment for a commercial or industrial site using only the site address, the facility type (for example, hospital, office building, manufacturing plant), and the facility size. Deriving these estimates with such minimal set of user-provided data is a significant technical challenge. DBO utilizes AWS to combine these user inputs with extensive datasets on energy usage, facility characteristics, and location-specific geographic information to generate more precise estimates than country-wide averages.

Users can then discover and choose from a combination of generation and storage technologies to decarbonize their facilities, such as solar panels, combined heat and power (CHP), thermal energy storage, battery storage, and more. The tool will select and size the optimal technology mix that makes the most sense for the user’s site, based on a complex optimization algorithm, that is designed to minimize the total lifecycle cost of employing these technologies, while still meeting the user’s specific decarbonization targets.

These customized decarbonization scenarios are based on an algorithm that is built on top of NREL’s REopt, which evaluates numerous technology combinations, sizes, and financial factors. This process is resource intensive and time consuming to run. By harnessing the scalable processing power and data storage of the AWS cloud, DBO is able to run these computationally intensive optimizations in the background. This process provides the user with the most cost-effective technology combination suggestion that is tailored to their needs.

Solution

The original DBO architecture utilized several AWS services such as Amazon Route 53 for domain name management, Amazon CloudFront to cache static assets, Amazon API Gateway to accept HTTP requests and trigger the backend application, AWS Lambda to connect the backend and run the ETL pipelines, Amazon Elastic Compute Cloud (EC2) to run the optimization algorithm, and Amazon Relational Database Service (RDS) to store data.

Figure 1: Original Architecture

To optimize the solution, the DBO team aimed to modernize their architecture, embracing a serverless approach to lower costs, reduce the application’s carbon footprint, reduce operational overhead, and simplify management. To do this, AWS now handles all the underlying infrastructure tasks like provisioning, scaling, patching, and maintenance. At the core of this modernization is a migration from their traditional RDS database to a hybrid data storage solution, leveraging Amazon Aurora Serverless for dynamic and relational data, Amazon DynamoDB for static, non-relational data (such as lookup tables and reference data), and Amazon S3 for raw, unprocessed data. This allowed DBO to avoid the overhead of managing and provisioning a traditional RDS instance, resulting in a 50% reduction in database costs. Additionally, the team integrated AWS Glue into their data pipelines, enabling the automation of the team’s data ingestion and transformation workflows, as well as the ability to handle larger data volumes more efficiently, when compared to custom AWS Lambda functions. The final piece of the modernized architecture consists of a serverless optimization process. Instead of running the solver in a traditional EC2 instance, the team leverages a combination of AWS Lambda, Amazon ECS, and Amazon SQS to create a process that triggers the optimization process only when it is necessary. Now, an AWS Lamba function triggers an ECS task, which runs the optimization solver and stores the raw results in an S3 bucket, allowing the DBO team to only pay for the compute resources they consume during the optimization process. This approach costs the team only a few cents per optimization instead of the estimated $4,000 annual cost of running an equivalent unoptimized EC2 instance.

Figure 2: Modernized Architecture

Looking forward

With the initial release of DBO, Siemens and AWS brought a simple, easy to navigate way for SMEs to optimize their buildings’ carbon footprint according to their specific targets. In the future, Siemens plans to build upon this success by adding in capabilities beyond single building carbon optimization. By continuing to leverage AWS services, DBO will include analysis across more than one building (campus analysis), a fleet analyzer to assist companies in transitioning to an EV fleet, and climate risk analysis to help identify a site’s ability to withstand extreme weather events. At the same time, Siemens and AWS will keep collaborating on ways to making the tool more sustainable and its architecture more efficient, flexible, and scalable.

Conclusion

By embracing a serverless-first mindset, SFS and AWS significantly optimized the costs, scalability, and environmental impact of DBO, while improving the overall efficiency and reliability of their data pipelines and optimization processes. The pressure to decarbonize and invest in sustainable practices continues to grow. Facilitating this directive begins with helping customers baseline their carbon footprint, understand how they can reach their goals, and imagine the potential to help deliver a sustainable future, while focusing on the sustainability and efficiency of the application itself. If you are in the US and interested in understanding your carbon footprint, please utilize the DBO tool and share any feedback. Together, we can multiply positive sustainable impact.

References:

Legal Disclaimer: THESE RESULTS ARE AN ESTIMATE ONLY. FOR YOUR SITE-SPECIFIC ANALYSIS WE ENCOURAGE YOU TO CONTACT AN ENERGY ENGINEER. SIEMENS FINANCIAL SERVICES, INC. (SFS) IS NOT A REGISTERED MUNICIPAL ADVISOR AND IS NOT AN ADVISOR TO OR FIDUCIARY OF YOU OR YOUR COMPANY. THESE RESULTS ARE PROVIDED “AS IS” AND SFS MAKES NO REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED, TO RECIPIENT, RECIPIENT’S CORPORATION, EMPLOYER, BUSINESS ENTITY OR ANY THIRD PARTY CONCERNING THE RESULTS, INCLUDING, BUT NOT LIMITED TO, WARRANTIES FOR PERFORMANCE, COMPLETENESS, ACCURACY, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. SIEMENS DISCLAIMS ALL RESPONSIBILITY FOR ERRORS OR OMISSIONS IN THE RESULTS, OR RECIPIENT’S OR ANY THIRD PARTY’S RELIANCE ON THE RESULTS.

AWS for Industries