In the background of China's goals of "Dual Carbon"，hydrogen，as clean energy，is expected to be widely used；however，it leads to large carbon emissions in the production process. Hydrocarbon Steam Reforming （HSR） is a common pro- cess for hydrogen production in refineries，in which accurate carbon accounting is hard to perform，due to the high fluctuation of feed- stock and low frequency of monitoring of key carbon emissions data. In this paper，the hydrogen production process of a refinery was taken as an example，and the HSR model was built by Aspen Plus to analyze the impact of different process parameters on carbon emis- sions in order to address the lack of accurate carbon measurements. Due to the large variation of some important parameters in the actu- al production process，it is hard for a steady model to produce sufficiently accurate simulation data . This performed experiment inter- connected MATLAB with Aspen Plus and changed model input parameters in accordance with the actual production data to achieve the dynamic simulation of the production process. The experimental results verified that the proposed model can well simulate the actual hydrogen production process and thus，provide data support for accurate carbon accounting.

To mitigate climate change by reducing CO₂ emission，carbon capture，utilization，and storage （CCUS） technologies have garnered significant attention. However，the required large scale investment and inflexibility of CCUS projects have greatly hindered their widespread applications. In this context，systematic source-sink matching has emerged as a key research focus，as scientific and efficient matching can optimize pipeline network design and reduce the overall costs of CCUS implementation. To this end，this study proposes a CCUS pipeline network layout optimization method using Density-Based Spatial Clustering of Applications with Noise （DBSCAN），offering a solution for CCUS pipeline network design. Firstly，the DBSCAN clustering is employed to cluster emission sources and storage sinks. Subsequently，a CCUS source-sink matching model is developed based on the minimum spanning tree method after comprehensively considering source-sink characteristics and cost components across operational phases to generate theoretical CCUS matching schemes. Finally，to address pipeline redundancy caused by multi-sources to one-sink configurations，an improved saving algorithm is applied to optimize the CCUS source-sink matching scheme. A hypothetical planning region is presented for testing of the proposed method，which demonstrates that the model does not only reduce deployment costs but also significantly shortens transportation distances. Compared with traditional methods，the total deployment costs decrease from 130 billion CNY to 98 billion CNY，by a reduction of approximately 24.6%，while the transportation distance is reduced from 4 075 km to 1 008 km，marking a decrease of 75.3%. These findings validate the adaptability and economic efficiency of the proposed method in complex CCUS scenarios，and the proposed method provides a feasible optimization pathway and theoretical foundation for CCUS system planning.