01 / FEATURED
Skip to main content
OUR BLOG
10 Oct 2024 | Saumya, Max & Kevin
12 MINUTES READ
Carbon removal is a unique market. Companies pay for a product they will never physically see. They probably don’t want a tonne of bio-oil or crushed limestone showing up at their doorstep. Historically, the voluntary carbon market has had misaligned incentives, allowing unscrupulous actors to exploit this opaque dynamic by selling low-quality (or, in some cases, outright fraudulent) credits.
We’re taking a different approach at Charm. We’ve partnered with Isometric who defines science-backed protocols and is responsible for reviewing every step of our process – for every removal – to ensure the credits that our customers are receiving are high-quality and permanent. Isometric partners with validators and verifiers that operate as the “on-the-ground eyes and ears” to ensure that claimed carbon removals are, in fact, happening in accordance with a protocol. Importantly, Isometric is paid directly by buyers (not Charm), which removes a major financial conflict of interest compared to how verification processes have been conducted in the past.
An apt comparison is to financial accounting. A registry defines protocols, which are the equivalent of GAAP Accounting Rules for a specific Carbon Dioxide Removal (CDR) method. A verifier acts as an auditor does in finance, reviewing all the work that has been done and ensuring accuracy and alignment to the accounting rules.
While this new approach we’re taking creates the right long-term incentives, it creates a new question for suppliers: how can a carbon removal supplier reliably monitor and report on removal operations that can be independently audited?
Charm has been working on answering that question for years, and today we’re excited to share the inner-workings of Ledger. You may have interacted with Charm Public Ledger to explore the net emissions that go into a Charm removal. Today we’ll be diving into the system we’ve built to collect, monitor, and report on net removals, conform to registry protocols, and measure and model emissions with the latest climate science.
There are many emerging approaches to digital MRV, and our hope is that others can use our learnings and work to shape their approach and avoid some common pitfalls as other suppliers start to deliver. After all, we want to see more high-quality carbon removal in the world!
First, before we get into the guts of Ledger, it’s important to understand a few key points about Charm’s approach to CDR. Charm uses biomass residues as a feedstock for fast pyrolysis, which produces a carbon-rich, low-energy-content bio-oil. From there, bio-oil is injected into EPA-regulated injection wells, where the bio-oil sinks and solidifies to be stored for millenia.
A simplified diagram of Charm Operations. Plants soak up the sun and CO2, which gets converted into bio-oil, which moves via railcar to a storage facility before being permanently stored underground
Charm’s vision is to operate a distributed fleet of modular, mobile pyrolyzers. We generate a small quantity of emissions at every step of our supply chain: transporting biomass to our pyrolysis site, operating pyrolyzers, and transporting and injecting bio-oil. We receive lab test results from carbon sampling. We account for the embodied emissions of all the materials and equipment we use, and equipment used by our partners. We also account for any counterfactual or indirect emissions impacts we make on existing uses of biomass, which is one of the most rigorous approaches in the industry.
Charm’s mobile pyrolyzers set up in the Inyo National Forest
To calculate our net carbon removals, we measure and model every step in our operations.
Ledger collects & models data across the Charm operation to ensure accurate carbon accounting
Let’s be real: most MRV starts with data tracked in a spreadsheet. That’s okay, and we are all about finding simple solutions that work.
But as our operation became more complex and we started putting tonnes underground, we realized we were outgrowing the spreadsheet and began to sketch the outline of Ledger with the following principles:
Accuracy & Conservatism: Our aim is to provide buyers with ironclad confidence that 1 Charm CDR credit equates to at least 1 tonne of permanently sequestered CO2e. This goal is embedded throughout Ledger - we round our emissions up and our removals down. We quantify uncertainty for all of our measurements (a legal-for-trade truck scale is only accurate to within 160 lbs, or the weight of the average human!) and propagate uncertainty to our net CDR calculation.
Transparency & Auditability: Every tonne of CO2e sequestered by Charm is verified by an independent verifier and made public on the Charm Ledger and Isometric Registry. The verifier is selected independently. The verifier reviews Charm’s operations with a fine-toothed comb and ensure that Charm can prove the accuracy & conservatism of our carbon accounting.The ledger serves as a centralized system to collect evidence for every measurement that’s used in our accounting.
Flexibility & Modularity: Charm is running a first-of-kind distributed physical operation with no shortage of unexpected developments. In Ledger, we use a generic data model that makes it easy to express all types of emission-generating activities. Ledger needs to be agnostic to where data is collected and stored, integrating with frontline operational systems and exposing its own data collection tools & APIs.
Ledger is architected around 3 core components:
Data Collection & Measurement: Raw data & samples are collected for MRV as bio-oil moves through Charm’s operation.
Emissions Modeling & Allocation: Data is processed into a set of models which track the physical inventory of bio-oil along with our process emissions
Reporting & Insights: Data and reports are assembled, reviewed and audited before being sent to Verifier for review. Ledger data is also used for generating insights about Charm’s operation & providing transparency to customers via the Charm Public Ledger.
This section walks through each of these components, some of the challenges we’ve encountered and what we’ve learned along the way.
Data required for MRV are highly diverse and generated through a variety of mechanisms. Ledger is designed to collect both structured data (for emissions modeling) and unstructured data (as supporting evidence of underlying operations). This includes everything from photos of truck scale tickets, equipment invoices, pyrolyzer sensor logs, structured forms from field teams, and API integrations with partners.
Example Truck Scale Reading. While lots of our operation is high tech, the world we operate in sometimes results to paper and pen
Collecting all this data is a combination of technology integration and human process. When we undertake any major change to our operations, we pull together a cross-functional team to review the protocol requirements and determine the datasets that we need to collect and how we will collect that data – before we operationalize the change. This is an important step to ensure that as our operations continue to evolve in accordance with the protocol and that we will be collecting the right datasets for verification.
The frontend interface for Ledger is built in Retool, which provides interfaces for data collection as well as analytics & insights to review automated data collection. Below you can see the interface for transferring oil from truck to tank. The origin and mass of the transfer is collected, allowing us to monitor the contents of any tank across the Charm operation.
Ledger interface for transferring from truck to storage tanks and preserving the origin of the oil (demo data)
One of the key feature requirements for Ledger is attribution - when calculating the net CDR for a given truckload of oil that we inject, we account for the emissions associated with the production and transportation of that specific batch of bio-oil.
Below you can see that data model in action—two tanker trucks full of bio-oil are transferred into a railcar (rail is much more cost- and carbon-efficient than trucking!). After the railcar trip, the oil is transferred into 2 new tanker trucks where it’s driven to the injection site and pumped down the hole.
Our teams use a growing number of frontline systems that are critical for MRV. Charm Operations uses Supervisory Control and Data Acquisition (SCADA) to manage pyrolysis, our finance team tracks invoices in Airbase, truck scale tickets are physically collected, railcar positioning updates come in via daily email. Ledger must keep track of the movement of bio-oil through this supply chain. Ledger backend computes emissions for each railcar and truck trip, and propagates emissions based on the physical movement of oil.
The ledger’s backend data model mirrors the structure of our physical operations - it’s a graph (a weighted directed graph, to be specific)! Nodes in the graph are Lots, or some distinct quantity of bio-oil. Edges are created when oil is transferred between physical containers. Edge weights are based on the weight of oil transferred.
Each Lot is associated with emissions - truck trips, railcar trips, etc. When we inject bio-oil we create a node with no outgoing transfers. We can attribute emissions to an injection by backtracking the graph and scaling emissions based on edge weights.
Tracking the movement of bio-oil like this requires considerable effort from our software and operations teams, but it’s essential for providing our buyers and verifiers with industry-leading transparency into the operations behind our CDR credits. Our approach is informed by other “high-stakes” supply chains where ensuring visibility and quality is critical, like food & agriculture and pharmaceutical manufacturing.
The final leg of the MRV journey is reporting to Isometric for verification.
This happens in two steps. First, data is transferred from Ledger data to the Isometric dMRV system via the Isometric API. This ensures that all the hard work that went into data collection and modeling is made available for the verifier to review and inspect our work. Isometric generates a calculation graph to visualize each removal.
The second piece is assembling the GHG Statement Report, which outlines operations for the reported period and details any anomalous data, considerations or divergences from the Project Design Description. The goal is to make this as minimal as possible, although we do document any data or operational changes that the verifier should be aware of in verifying removals.
We have had a ton(ne!) of learnings along the way in building and operationalizing Ledger. While we are still in the early phases of scaling up our operation, we’ve already learned so much. A few of our key learnings:
MRV is a distributed responsibility – everyone has a role to play in ensuring verifiability of removals. Each step in the process of producing and injecting bio-oil produces data points that must be captured and used to produce accurate, verifiable carbon removal claims.
Our organizational structure reflects this – we have an MRV team that focuses on protocol development, project validation, system integration and enabling the verification process. Data collection is owned by frontline teams. The datapoints that need to be captured are made clear to all operators and supervisors, documented in Standard Operating Procedures (SOPs), and reviewed for completeness. It’s everyone’s job to make sure we have the data and tracking in place to verify our removals.
There are many ways to implement the digital and data side of MRV. You can start in a spreadsheet, buy a vertical-specific MRV software, or extend your Enterprise Resource Planning (ERP) solution to support MRV. The tech needs to support the frontline operators, and those operators need to be trained and equipped to collect the data required for MRV.
Data collection and quality control are baked into all of Charm’s operational SOPs. Just as operators are trained to run a machine, they are trained to collect the operational data about its function that is required for MRV.
One important learning is that there are often ways to simplify your MRV data collection if you think about the minimum data required to verify a removal.
One example we’ve encountered is in tracking total bio-oil production. For example, our operations and process engineering teams may want to collect an estimate of bio-oil production for a given pyrolyzer every minute or hour for system improvements. That may sound like something we’d want for MRV, but the requirements are a little different. Instead, we need to measure total bio-oil production with less granularity (i.e. daily output is sufficient) but with extreme precision (i.e. measured on calibrated scales).
Rather than trying to get hyper-granular and build complex automations, it can be helpful to step back and think about the requirements for MRV as standalone and come up with simple solutions to collect that data.
As we look ahead, we see a lot of opportunities to extend the core MRV functionality we’ve developed in Ledger.
First, because Ledger is such a rich repository of data, there are opportunities to extend the insights and analytics capabilities to inform production forecasting, process improvements and even our growth model.
As Charm expands the markets that we’re in, we know that there will be a diverse set of reporting integrations on the horizon. The Department of Energy, state compliance markets, reporting platforms like CDR.fyi, and other regulated markets each different requirements and considerations, and so the ability to flexibly connect to new APIs and generate custom reports is something we’re exploring.
Finally, we plan to expand on the transparency we provide to Charm customers. We work with a variety of partners we looking forward to featuring who are helping Charm scale, from sustainable energy to biomass sourcing and transportation. There are a lot of fascinating photos and videos we collect that will provide customers with a front row seat to carbon removal operations and we plan to expand beyond structured data to help everyone peek inside Charm’s operation.
Ideas for how we can improve Ledger, or want to chat MRV? Get in touch!
Find it interesting? Share!
Saumya Jain
Senior Software Engineer, Staff Scientist
Max Lavine
Carbon Protocol and Verification Lead
Kevin Niparko
Head of Product & MRV
Subscribe to follow our journey to inject bio-oil into deep-geological formations, Charm permanently puts CO2 back underground.
Find it interesting? Share!
Over the past year, we completely redesigned our approach to biomass processing. Our goal was to minimize our carbon footprint while maximizing our throughput. The journey to the “perfect chip” took us across the country to test wood chippers and loading systems; it required hundreds of engineering hours spent analyzing chipping yields and updating test plans for drying techniques; and it ushered in a new fleet of processing equipment.
Mark Hamann
Sr. Mechanical Engineer
Over the past year, we completely redesigned our approach to biomass processing. Our goal was to minimize our carbon footprint while maximizing our throughput. The journey to the “perfect chip” took us across the country to test wood chippers and loading systems; it required hundreds of engineering hours spent analyzing chipping yields and updating test plans for drying techniques; and it ushered in a new fleet of processing equipment.
Humanity has emitted hundreds of gigatonnes of CO₂. Now you can put it back underground.