Reducing Gas-to-Power Emissions for Data Centers with Pipeline-Connected Carbon Capture
TAMÁS UNGÁR, Capsol Technologies, Oslo, Norway
Data centers are becoming one of the largest and fastest-growing consumers of electricity worldwide, driven by the rapid expansion of cloud computing and artificial intelligence (AI).1 Data center operations require a continuous supply of high-quality and highly reliable power, leading to fixed electricity demand (FIG. 1).
Consequently, new data centers are developed in regions that offer available land, access to cooling water, favorable permitting conditions and reliable gas infrastructure, such as the U.S. Gulf Coast and parts of the Middle East. In many cases, grid expansion cannot keep pace with rising demand, pushing developers to rely on nearby power generation. While renewable energy sources play an important role in the energy mix, they cannot guarantee constant power availability, making gas turbines a key solution, especially where fast deployment is required.
At the same time, corporate climate targets and regulations demand notable reductions in emissions from gas-fired power generation. This creates a central challenge: how to maintain the reliability of gas-to-power systems while reducing their carbon footprint in an efficient and scalable manner.
Gas Pipelines
Existing gas pipeline infrastructure represents a strategic advantage for supplying power to data centers. These networks are already broadly developed and operate with high availability, providing a reliable and continuous fuel supply without the long lead times associated with new energy infrastructure. Access to pipeline gas enables power generation to be located close to data centers, reducing transmission losses and minimizing dependence on large-scale grid expansions. As a result, gas-fired power plants can often be deployed significantly faster than major grid reinforcement projects, improving project timelines and reducing overall development risk. By utilizing existing pipeline infrastructure, developers can enhance the economic and technical viability of data center projects while maintaining the high reliability required for continuous operation.
Decarbonization
Decarbonizing gas-fired power generation for data center applications presents several technical challenges that are closely linked to the operating characteristics of both the power plant and the capture technology. Most carbon capture systems are designed for steady baseload operation and can tolerate only limited load variation. This operating profile aligns with data center power demand, which is normally stable and continuous, allowing capture systems to be sized and optimized for near-constant operation.
However, gas turbines produce high-temperature exhaust gases, whereas conventional solvent-based post-combustion capture systems are designed and therefore able to handle significantly lower flue gas temperatures. Cooling the exhaust gas to acceptable levels introduces additional energy losses and increases operating costs.
An additional challenge lies in thermal integration: many capture technologies require external heat or steam for solvent regeneration, requiring auxiliary boilers, increased water consumption and added system complexity. In regions with limited water availability, these requirements can become a critical constraint, highlighting the need for capture solutions that enable efficient thermal integration and minimal additional resource use.
High-Temperature Concepts
One approach to overcome the challenges associated with high-temperature gas turbine exhaust is to design carbon capture systems that can operate at elevated temperatures, rather than forcing the exhaust gas to conform to low-temperature capture processes. High-temperature capture concepts, such as hot potassium carbonate-based systems, are well suited to this application, as they can directly utilize the thermal energy available in gas turbine exhaust streamsa.
As hot potassium carbonate capture requires flue gas compression, the processa incorporates a turboexpander to recover energy from the compressed gas. This configuration allows a substantial share of the compression work to be recovered and allows net power generation after supplying the energy required for the capture system and carbon dioxide (CO2) conditioning. Efficient internal heat recovery and thermal integration eliminate the need for auxiliary boilers or external steam, reducing system complexity, water consumption and overall efficiency losses.
The process also tolerates oxygen and is suited to the relatively clean exhaust gases of gas turbines. While load flexibility is limited by rotating equipment, this constraint aligns with data center power plants, which typically operate at or near baseload, making high-temperature capture viable for this application (FIG. 2).
Typical Project Configurations
Industrial size gas turbines in the 30 megawatt (MW)–70 MW range are commonly deployed for individual data centers due to their modular design, standardized layouts and relatively short construction timelines. These units enable fast project execution and phased capacity additions, making them suitable for the rapid development cycles of data center projects. Their size also allows power generation to be located close to where it is needed, reducing transmission requirements and improving overall system efficiency.
Larger gas turbines or turbine clusters, which may serve the largest data center projects as well as large scale utility supply, produce higher CO2 volumes that improve the economics of carbon capture through economies of scale. Such configurations are attractive in regions with access to shared CO2 infrastructure, including regional pipeline networks and centralized storage hubs. In some locations, enhanced oil recovery (EOR) sites can further improve project economics by providing near-term CO2 utilization options.
Hybrid configurations are also emerging in which multiple industrial size gas turbines feed a shared carbon capture system. These arrangements fit campus-style data centers and industrial parks, where power demand increases over time. Capture capacity can be expanded incrementally alongside generation assets, enabling scalable decarbonization while maintaining high reliability and minimizing upfront capital expenditure (CAPEX).
CO2 Transport and Pipeline
Capturing CO2 is only part of the value chain. In addition to generation and capture, CO2 must also be conditioned before it is transported or otherwise utilized. This includes dehydration, removal of oxygen and compressing the gas to pipeline pressure. The total quality specifications depend on the transport network and storage site.
For regions with developing CO2 hubs, requirements to blend CO2 from various sources may apply. Therefore, proximity to storage sites, such as depleted gas fields, plays a role in project viability. This mirrors the gas supply side: pipelines bring the fuel in, and pipelines take the CO2 out (FIG. 3).
Economics
Reducing CAPEX and accelerating the speed to market are key requirements for deploying carbon capture in gas-to-power systems serving data centers. A primary CAPEX driver is the level of integration between the power plant and the capture system and the ability to utilize available waste heat for solvent regeneration. Capture concepts that efficiently recover exhaust heat reduce or eliminate the need for auxiliary boilers, steam systems and additional balance-of-plant equipment, resulting in simpler plant layouts, lower installed costs and shorter construction timeline.
Designs based on proven chemistry and standard industrial equipment further improve cost predictability and bankability, which is especially important for first-of-a-kind projects. Modular and pre-fabricated capture systems play a critical role in shortening deployment by shifting construction activities from site to factory, reducing onsite labor and minimizing interface risks. Standardized modules enable parallel engineering and manufacturing, support repeatable designs and can simplify permitting processes. Together, these factors reduce execution risk, improve schedule confidence and create a scalable pathway for rapid replication as data center capacity continues to expand.
Decarbonized Gas-to-Power Networks
Data centers will continue to depend on gas-fired power for reliability. Currently, no single alternative exists that could fully replace it at scale and as rapidly as gas turbines can be deployed. Existing gas pipelines offer a low-cost and flexible way to supply energy and will remain so for years to come.
Decarbonizing gas turbines requires capture systems and solutions that are fit to purpose, meaning matching the capture process to the gas turbine exhaust conditions is crucial. Efficient waste heat utilization provides an efficient and practical pathway, especially when combined with modularization and access to CO2 transport and storage.
Storage availability, policy frameworks and regulations will determine how fast these solutions are installed. From an engineering and industry perspective, the tools already exist to deploy and reduce emissions from gas-to-power systems.
NOTE
a The CapsolGT® process
LITERATURE CITED
1 Goldman Sachs Research, “Data center power demand: The 6 Ps driving growth and constraints,” November 10, 2025, online: https://www.goldmansachs.com/insights/goldman-sachs-research/data-center-power-demand-the-6-ps-driving-growth-and-constraints
2 BNP Paribas, “The future of carbon capture and storage: Strategies and challenges,” April 14, 2025, online: https://cib.bnpparibas/the-future-of-carbon-capture-and-storage-strategies-and-challenges/
About the Author
TAMÁS UNGÁR is a Senior Engineer with more than 10 yrs of experience in the power generation sector, primarily focused on commissioning activities for greenfield natural gas-fired power generation projects. His background includes mechanical and process engineering roles on complex international projects. Ungár earned an MS degree in aerospace engineering and has been working in carbon capture at Capsol Technologies for 3 yrs, contributing to feasibility and pre-FEED studies, technology assessments and project development from concept through advanced development stages.