PipeSence demonstrated the technology’s capability on a South Texas pipeline
PipeSense, a midstream technology provider, is expanding its support to emerging CO2 applications and networks
By successfully designing, implementing, and testing a new dynamic pressure analysis-based leak detection system, PipeSence demonstrated the technology’s capability on a South Texas pipeline transporting dense-phase and supercritical carbon dioxide (sCO₂) at 2,000 psi. Deploying its team on-site for testing, PipeSense was responsible for a 12-mile stretch from a compressor station to a client’s handoff point in the state.
Applying a Dynamic Pressure Analysis (DPA) method enabled the leak detection system to use two Field Processing Units (FPUs) equipped with high-sensitivity pressure sensors, one at each end of the pipeline segment.
Installation commenced in Q3 2025, while tests continued throughout Q4 2025.
Upon completion of the project, the system capability can detect leak events within 2-3 minutes and with location precision down to <20 ft. Over the five-month period, one false positive and zero downtime were observed, with the system achieving sensitive, accurate detection and addressing pipeline network challenges, including speed-of-sound variability, unplanned compressor station shutdowns, and changes between dense and supercritical phases.
Demonstrating the efficacy of a Machine Learning-enhanced DPA system for leak detection in sCO₂ pipelines, the project has the potential to mark a crucial milestone in supporting the growing demand for Carbon Capture and Storage (CCS) infrastructure. With US CO₂ pipelines expected to grow 20x by 2050, the availability of PipeSense’s technology will enhance safety and minimise environmental risks from undetected leaks.
PipeSense is also planning for further testing and implementation within CO₂ networks and infrastructure in the US and globally.
Stuart Mitchell, president and CTO of PipeSense, said, “Supercritical CO₂ pipelines operate near the critical point, where fluid properties such as density and compressibility exhibit highly nonlinear and rapid variations. Standard Real-Time Transient Models (RTTM) and statistical models, which rely on simplified equations of state and linear assumptions, struggle to accurately predict pressure and flow transients during operational fluctuations. These modelling inaccuracies produce persistent discrepancies that appear as false leak signatures when residuals or pressure anomalies are misinterpreted. Our dynamic pressure analysis method offers a superior, mass-flow-independent alternative, paving the way for reliable monitoring in expanding CCS infrastructure.
The simplicity of our system was on full display during this recent implementation. Through just two FPU-connected pressure sensors and a flexible dashboard interface, the system demonstrated its effectiveness and highlighted its potential to scale to larger CCS networks, many of which will be required by the expected growth in CO2 pipelines.”
