The Problems with Carbon Accounting in 2024

Carbon accounting is a complex process.

You need to identify organizational and operational boundaries, select and consistently apply an appropriate consolidation approach and continually identify relevant emission sources across your organization. You then need to identify and collaborate with data owners, brief data owners on data requirements, collect activity data, quality assure, process and normalize that data and provide feedback to data owners on data issues. Finally, you can source and apply the most appropriate emission factors to calculate emissions results and model unavoidable data gaps. 

Any misstep increases the risk of misreporting, leaving you open to reputational damage, fines or even legal action. 

However, there is a common thread to most of these challenges: data management workflow. 

Ultimately, carbon accounting is a data collection and analysis task, reliant on collaboration between many individuals. The central issue is a lack of suitable technology able to guide and automate the entire process for all parties. This leaves organizations heavily reliant on human labor to fill in the blanks — chase records via email and manually inputting data in spreadsheets — a time-consuming, error-prone, and costly undertaking. 

But it doesn’t have to be this way. 

Only 10% of companies measure greenhouse gas (GHG) emissions comprehensively. This figure balloons by 250% for organizations that automate part of their carbon accounting process.¹

What if you could automate the whole process?

Here, we’re going to explore why data management workflow is the central problem with carbon accounting and what you can do to free up resources and avoid the risk of misreporting. 

‍

The problem: Carbon accounting data management 

Your main carbon accounting challenge might be collaboration between teams, maximizing data coverage, understanding carbon accounting standards or reporting requirements, or any number of secondary issues. But if we break down the challenge of data management, you can see how it’s all related. 

Challenge 1: Discovering data

You face three issues when scoping out the data required for accurate carbon accounting.

1. Identifying the ideal activity data you need for complete carbon accounts.
2. Identifying who owns that data in your organization, and how much of the ideal data they do and don’t have.
3. Coordinating with third parties (e.g. building managers) to get the data you don’t have within your organization.

Organizations also need to determine which sources of emissions to include and exclude, based on whether your organization has operational or financial control over the emissive asset. Establishing a boundary and reporting against it consistently is complex. 

Different departments, or physical locations, may manage their own data collection or obstruct collaboration. The larger your organization and the more locations you have, the more time-consuming and tedious this task becomes. 

Challenge 2: Collecting and processing data  

The activity data across your organization won’t all be in one format. Some records will be semi-structured, some not. The result generally means that someone, often many people — in-house, or a third party — will need to manually comb through all those primary records and transcribe the relevant data points such as dates, activity type, amounts, and units, into a standardized format. More than any other carbon accounting step, this task is time-consuming and error-prone, and it usually breaks the audit trail. 

The additional risk you face is poor engagement from across your organization in the carbon accounting process. Laborious and time-consuming manual data entry tasks can make people feel like you’re wasting their time. This can result in poor data contributions, which heightens the risk of misreporting. It also drives up the direct and indirect costs of carbon accounting.    

Challenge 3: Estimating data gaps

It’s not always possible to gather a complete set of records on all activities contributing to your organization’s emissions profile. To avoid the risk of under-reporting, it’s important to reliably identify gaps, and make accurate estimates to fill them. This requires selecting and applying the most appropriate modeling approach based on the data that is available. This can include extrapolation or intensity-based estimation, which may also need to account for seasonality. The tedious process of manually identifying data gaps and applying estimations often leaves this step squeezed in at the last minute, or unfinished, contributing to erroneous results. 

Challenge 4: Maintaining & auditing data 

Once your data has been discovered, collected, and processed, it’s important to maintain those original records in case they need to be audited. With primary and secondary records split across multiple documents, platforms and spreadsheets, it’s easy for those sources to fall out of date, get lost or forgotten about. 

Maintaining accurate records is also important for rebaselining calculations to make meaningful comparisons of emission data over time. This is an important part of GHG Protocol standards. ²

‍

The game changer: Generative AI

There are tools and resources that solve some of these carbon accounting challenges. But 99% of solutions still rely on manual data extraction from primary sources using human labor, and some form of spreadsheet. This fails to address the most time-consuming and error-prone part of the carbon accounting process, breaks the audit trail, and interrupts the seamless end-to-end workflow.  

What generative AI allows you to do is deal directly with your organization’s primary data records — regardless of their structure, file type, and language — and automatically extract relevant information for carbon accounting. Partnered with a simple end-to-end data collection workflow, this provides an incredibly simple solution for getting emissions results with unparalleled auditability. 

Traditional carbon accounting software is useful for project management and basic data management, but without automatic data extraction, it can’t solve the central challenge of carbon accounting.  

‍

The solution: End-to-end workflow automation  

If you can use generative AI to solve carbon data extraction, it’s possible to create a tool that spans the entire carbon accounting workflow within a single platform. 

This observation is where Ideagen Carbon Accounting started. Including data extraction, we identified four key capabilities needed to re-imagine carbon accounting.

1. Guidance on requirements: A smart discovery system that uncovers all emissions sources you need to report on, the ideal data required, and the data actually available within your organization.  
2. Engagement tools: The ability to easily identify and invite the people who have access to the required data to submit that data. The ability for data contributors to instantly know whether they have provided all the required data, and whether there are any issues with it, without anyone else needing to review it manually. 
3. Automated extraction from primary records: The application of generative AI to automatically and directly scan primary data records to extract and use the relevant information needed for carbon accounting. 
4. Long-term data storage and automated estimates: A central location to store all of your primary records and past reports to ensure rapid audits, and the ability to automatically and transparently model any unavoidable data gaps to avoid the risk of underreporting.  

If approached independently, each of these steps adds complexity, requiring you to use a patchwork of human intervention and software support. When merged into a single platform, you get a solution that is more than the sum of its parts. 

• All data discovery feedback loops occur instantaneously, and within the same platform the data is stored  — ensuring rapid results, low-touch management, and total visibility. 

• Unavoidable data gaps are identified on-the-spot, automatically modeled, and properly evidenced  — eliminating blind spots, and time delays due to human intervention.

• All primary records are stored in the same tool used to produce the emissions results — providing instant access for review, audits and rebaselining. 

Complexity will hold you back from credible reporting and is a big barrier to entry. If you simplify how you discover, collect, and process data, you remove the most complex part of the process. 

The result is a transformative approach to carbon accounting. AI applied to carbon accounting cuts days of work into minutes while ensuring total defensibility.

‍

_{1 Why Some Companies Are Ahead in the Race to Net Zero}

_{2 https://ghgprotocol.org/sites/default/files/2022-12/Base%20Year%20Adjustments.pdf}

‍