The Starting Problem Is a Missing Connection
An industrial site may produce a CO2 stream and have a reason to reduce the amount released. That does not mean it has a complete route to long-term storage. Equipment, transport, geological access, permissions and commercial responsibility still have to fit together.
The problem resembles an infrastructure-coordination task more than a simple sale of a device. A capture unit without a destination can be stranded; a storage site without ready customers can be underused. The useful offer is the connection between them.
That framing explains the significance of a network. It can coordinate several sources and destinations, potentially sharing infrastructure. But the value appears only where the physical chain and economic agreements actually work.
The Industrial Customer Has More Than One Requirement
An ammonia producer, steel-related facility or gas processor must keep its main production system functioning. A CO2 project needs to integrate with that system without assuming the customer's entire plant can be redesigned without cost.
The stream also has characteristics that affect handling. Preparation, dehydration or compression may be needed before transport. Responsibilities must be allocated between the customer and service providers so that the interface can operate reliably.
CF Industries' Donaldsonville project provides a concrete example. The customer built a dehydration and compression unit, while ExxonMobil undertook transport and storage. The operating chain therefore depended on both sides, not solely on the existence of a pipeline. [S54]
Existing Experience Helped, but Did Not Complete the New Offer
ExxonMobil already had subsurface, processing and project capabilities. Those skills are relevant to understanding storage and building industrial systems. They did not automatically create a ready commercial service for every external emitter.
Third-party service adds contractual and coordination work. The parties need to agree on what is supplied, who prepares it, where it goes, how performance is measured and how economic rights are allocated.
The distinction matters because familiar technology can still support a new business model. The challenge is assembling the components into a dependable customer proposition. Experience reduces some uncertainty, but does not remove the work needed at every new connection.
Denbury Changed the Infrastructure Starting Point
The acquisition of Denbury completed in November 2023. It added a CO2 pipeline network, producing assets and people with operating experience. That gave ExxonMobil an existing platform rather than requiring every route to be built from nothing. [S11]
The network's historical use was closely connected with enhanced oil recovery. It was not originally a complete set of dedicated storage services for all nearby industrial facilities. The acquired business also produced oil and gas, which remained part of the transaction's economic scope.
The strategic opportunity was to connect that platform with additional sources and future destinations. Ownership transfer created the ability to pursue the transformation; it did not prove the transformation was finished at closing.
Pipeline Mileage Is Not the Same as a Working Customer Route
A network's length indicates physical reach, but not every potential customer's ability to connect. Route, available capacity, stream specifications and the operating destination determine whether a particular connection is feasible.
Additional facilities may be required, and a customer's commissioning schedule may differ from the transport or storage schedule. A map can make the system look complete before those interfaces are resolved.
The value of infrastructure is therefore relational. A route becomes commercially useful when it connects a ready source with a usable destination under workable terms. Counting all nearby emissions as immediate demand would skip the engineering, permission and investment decisions that make the route real.
A Storage Opportunity Needs Its Own Evidence
Geological potential is not the same as an operating injection site. A project needs characterization, design, permissions, construction and continuing controls. The suitability of one location cannot be inferred merely from regional experience with oil and gas.
EPA's Class VI framework addresses wells intended for long-term geological sequestration, with requirements around construction, operation, monitoring and protection of underground sources of drinking water. The framework defines a regulatory and technical system, not a blanket endorsement of every proposed project. [S56]
That distinction creates another sequence to manage. Customer capture readiness and pipeline access may advance before a dedicated destination reaches operation. The network developer must either align the timing or provide a suitable alternative route.
EOR Provided an Interim Path
CF's startup announcement describes initial storage through enhanced oil recovery, with a planned later transition to dedicated storage. The February 2026 NG3 account describes the same broad interim arrangement for the first two operating customer projects. [S54][S70]
This is a material part of the story. The initial commercial service did not wait for every proposed dedicated Class VI facility, but it also should not be described as already using those future sites.
EPA distinguishes Class II oil-and-gas-related wells, including enhanced recovery, from Class VI sequestration wells. The physical experience overlaps in some respects, while the purpose and applicable regulatory framework remain distinct. Treating all CO2 injection as one undifferentiated activity would conceal that transition. [S57]
The First Customer Startup Made the Chain Concrete
CF Industries announced the Donaldsonville startup in July 2025. The project could enable transport and storage of up to two million metric tons of CO2 annually. That statement describes a capacity or intended operating scale, not a measured full-year total captured at the startup date. [S54]
The milestone nevertheless changed the evidence. There was now an operating customer connection rather than only a development announcement. The chain joined a specific industrial source, preparation equipment and an actual transport/storage arrangement.
That is the commercially useful progression: identify demand, prepare the interfaces, commission the system and then establish repeatable operation. Each stage supplies evidence that the earlier concept alone could not provide.
NG3 Added a Different Source
In February 2026 ExxonMobil announced that it was transporting and storing captured CO2 from the New Generation Gas Gathering project in Gillis, Louisiana. The customer system gathers and treats gas, removing CO2 before the gas returns to Gulf Coast markets. [S70]
The source is therefore not identical to ammonia production. Different industrial processes can supply streams with different preparation and operating requirements, even when they use parts of a common transport/storage network.
The announcement described up to 1.2 million metric tons per year expected to be removed. That is not an annual measured result or proof that all gas eventually supplied through related markets is emission-free. The service addresses a defined part of a larger lifecycle.
Nucor Extended the Operating Customer Base
The September 2026 Nucor startup added a direct-reduced-iron facility in Convent, Louisiana. ExxonMobil described it as the third active third-party customer operation and the second startup of 2026. [S19]
That progression supports the claim that the network was expanding through actual connections. It does not establish that every contracted project was operating or that the maximum annual quantity was already flowing continuously.
The customer context matters as well. Reducing a defined process-emission stream can change the characteristics of an industrial product without eliminating every emission associated with inputs, energy or downstream use. The milestone should be stated at its actual boundary rather than expanded into an unrestricted claim about all steel.
The Payment Model Was Not One Standard Disposal Fee
Nucor's filing disclosed that ExxonMobil would receive relevant tax credits and pay Nucor a fee for each ton of CO2 supplied. The emitter therefore was not simply paying an identical disposal charge assumed from a generic service analogy. [S55]
CF's startup announcement described its own expected qualification for Section 45Q credits. The commercial allocation differed, demonstrating why the rights and obligations must be read project by project. [S54]
These arrangements can combine industrial value, infrastructure economics and incentives. None of the disclosed gross quantities alone establishes provider profit. Capital, operating costs, monitoring and the contractual allocation of benefits remain part of the business calculation.
Rose's Permit Was a Later, Distinct Milestone
In September 2026 the Texas Railroad Commission approved the Rose permit by a 2–1 vote. That later action changed the regulatory stage from earlier draft materials. The dissent is part of the record, not a judicial finding that an operating project was unsafe. [S71]
ExxonMobil's contemporaneous account still describes Rose as moving toward operation. Approval therefore should not be translated into a verified annual injection result or a claim that every customer's CO2 had already moved to that destination. [S19]
The sequence matters because the network can contain assets at different stages simultaneously: operating customer connections, existing transport and a dedicated storage project advancing toward use. One status label cannot accurately describe all of them.
Monitoring Approval Has a Different Purpose
EPA's Subpart RR record lists the Rose monitoring, reporting and verification decision in April 2026. That is a defined reporting approval, separate from the later injection-permit chronology. [S72]
EPA explicitly says a Subpart RR decision does not replace Underground Injection Control obligations or constitute an endorsement of the project or parties. It also distinguishes its reporting role from the IRS-administered tax-credit system.
This prevents a common shortcut: one approved document is not proof that every regulatory, technical and commercial condition is complete. The project needs the relevant permissions and controls for the activity actually being performed, and operating claims need measured evidence after the approvals.
The Service Creates Long-Term Obligations
Operation is not the final step. Storage systems require monitoring, reporting and eventual closure work. Financial and technical responsibility must persist beyond the first injection and the initial customer announcement.
EPA's reporting framework uses defined data and a mass-balance approach for covered facilities. The quantities received, injected, produced, leaked and sequestered have different roles in establishing the result. [S72]
Those obligations are part of the economics. A business description that counts only the initial transport volume or credit opportunity would omit continuing costs and responsibility. The durability of the service depends on managing the full operating life, not merely reaching a ribbon-cutting milestone.
Stored CO2 Is Not Automatically Net Emissions Avoided
The environmental outcome needs an appropriate counterfactual and boundary. Energy used by capture, transport and injection matters, as do uncaptured emissions, leakage and the broader industrial system. A capacity number is not a lifecycle assessment. [S65]
EOR adds another reason to specify the accounting boundary because the activity is connected with oil production. Recognizing that connection does not require assuming every project has the same outcome; it requires examining what the claim actually includes.
The useful conclusion follows measured performance. A project can provide a real storage service while leaving other emissions outside its scope. Reducing one stream is meaningful without being described as eliminating the customer's entire footprint or making every related product climate-neutral.
The Parent Business Still Has to Earn Its Place
For ExxonMobil, the network uses familiar capabilities in a different customer proposition. The overlap can make development more practical, but does not guarantee an acceptable return or remove the dependence on market and policy conditions.
The annual reporting framework keeps Low Carbon Solutions financial activity within Corporate & Financing pending materiality conditions. It is not a fifth separately disclosed reportable segment whose profit can be inferred from the number of contracts. [S33]
The current carbon-management service therefore needs its own stage and responsibility description. Corporate financial scale can support the undertaking, but cannot substitute for qualified demand, actual flow and continuing performance.
The Result Is a Developing Commercial System, Not a Completed Promise
The chronology shows real progress: infrastructure acquired in 2023, CF startup in 2025, additional customer operations in 2026 and a dedicated-storage permit advancing the next part of the system. It also shows different stages coexisting rather than one instant transformation.
The central business lesson is coordination. Industrial sources, preparation equipment, transport, storage, commercial rights and monitoring have to work together. A strength in one does not eliminate a weakness in another.
The continuing test is whether more connections become dependable, measured and economically sustainable operation. That is a stronger and more precise claim than either dismissing the entire business as only a plan or treating all contracted future capacity as completed emissions reduction.