Petroleum refineries are among the most energy- and hydrogen-intensive industrial systems and play a central role in supplying fuels and chemical feedstocks. Reducing their emissions is therefore an important component of climate strategies. However, existing studies commonly rely on technology-by-technology assessments or static system configurations, implicitly treating decarbonization as the additive deployment of independent options. Such approaches are inadequate for refineries, where material, heat, power, and hydrogen interactions constrain feasibility, shape technology performance, and create path dependence over time.

This thesis develops an integrated modeling and optimization framework to analyze industrial decarbonization from a system integration and transition planning perspective. Refining processes and mitigation technologies are formulated as process units with standardized material, heat, power, and hydrogen interfaces under harmonized thermodynamic and techno-economic assumptions. This representation enables heterogeneous technologies to be consistently integrated and compared within a unified system boundary, allowing interactions and feasibility constraints to emerge endogenously.

Building on this framework, decarbonization is formulated as a multi-period decision problem rather than a static technology selection exercise. A mixed-integer linear programming formulation is used to generate integrated system configurations and to construct alternative investment sequences under evolving external conditions. Future uncertainty is represented through systematic generation of market trajectories, and the resulting transition pathways are evaluated using harmonized performance indicators and a Monte Carloâ based robustness assessment. This approach enables explicit assessment of sequencing effects, technology lock-in, and trade-offs between cost, emissions, and system flexibility.

The framework is applied to a representative refinery to analyze scope 1 and scope 2 emission mitigation strategies under diverse techno-economic conditions. The results show that energy efficiency and heat integration consistently emerge as the first and lowest-cost levers, while the competitiveness of electrification, hydrogen supply options, and carbon capture technologies depends strongly on system-level interactions. Extensions to changes in product demand demonstrate that shifts in refinery outputs significantly affect hydrogen intensity, heat integration patterns, and preferred mitigation strategies.

While the present application focuses on emissions within refinery boundaries, the proposed framework is not limited to scope 1 and scope 2. Its generalized representation of material and energy flows, combined with multi-period decision modeling under uncertainty, provides a foundation for analyzing broader system transformations, including demand-side changes and scope 3 emissions.

Rather than prescribing a single optimal solution, this thesis provides a decision-oriented framework for exploring feasible, robust, and temporally consistent transition strategies in complex industrial systems.