{"id":{"repo_id":"mcmaster","oai_identifier":"oai:macsphere.mcmaster.ca:11375/33566"},"canonical_url":"https://search.ndltd.org/etd/mcmaster/oai:macsphere.mcmaster.ca:11375/33566","repository":{"repo_id":"mcmaster","name":"McMaster University","base_url":"https://macsphere.mcmaster.ca/server/oai/request"},"display":{"title":"Adaptive and Uncertainty-Aware Optimization of Green Hydrogen Production via Grid-Connected PEM Electrolyzer Systems","abstract":"Green hydrogen produced through proton exchange membrane (PEM) electrolysis can support decarbonization, but its economic and environmental performance depends on interacting factors including electricity prices, the timing of electricity consumption, renewable availability, and grid carbon intensity. This thesis develops an adaptive and uncertainty-aware scheduling framework for a grid-connected multi-electrolyzer plant with hydrogen compression, storage, and on-site renewable generation. The objective maximizes profit while accounting for electricity costs, cold starts, and unmet demand, with terminal inventory and carbon penalties included where applicable. First, the scheduling problem is studied using grid electricity only, with electricity price as the main uncertain input. A deterministic Single-Execution framework provides the baseline, while rolling-horizon optimization (RHO) updates electricity-price forecasts and re-optimizes as operation proceeds. Across the grid-only case studies, RHO achieved higher realized profit on average than Single-Execution, demonstrating the potential value of incorporating updated information into scheduling decisions. Building on this foundation, the framework is extended to include on-site wind and solar generation, renewable forecasting, and hydrogen carbon intensity (CI). This allows electricity prices, renewable availability, and grid CI to be considered together when determining plant operation. Renewable integration reduced grid dependence and improved the ability of the plant to satisfy hydrogen-CI limits, particularly under high grid-CI conditions. Forecast updating also improved performance when new information led to better implemented decisions, although repeated re-optimization did not always result in a better overall schedule. Finally, two-stage stochastic programming is incorporated within RHO to form RHO+TSP, allowing multiple possible electricity-price, wind, and solar trajectories to be considered within each optimization. Single-Execution, RHO, and RHO+TSP are then compared under realized conditions and across a common external bank of 1,000 scenarios. Single-Execution and RHO both performed strongly, while RHO+TSP performed well in some cases but was less consistent and substantially more computationally expensive. RHO provided the most consistent balance among adaptability, economic performance, operational reliability, and computational effort. Overall, this work provides a modular framework for examining how forecast updating, renewable availability, CI constraints, hydrogen storage, realized renewable deviations, and explicit uncertainty interact in grid-connected green hydrogen scheduling. Future work could incorporate realized renewable deviations directly into the rolling optimization framework, evaluate longer operating periods, and explore integrated design and scheduling under uncertainty.","abstract_html":"Green hydrogen produced through proton exchange membrane (PEM) electrolysis can support decarbonization, but its economic and environmental performance depends on interacting factors including electricity prices, the timing of electricity consumption, renewable availability, and grid carbon intensity. This thesis develops an adaptive and uncertainty-aware scheduling framework for a grid-connected multi-electrolyzer plant with hydrogen compression, storage, and on-site renewable generation. The objective maximizes profit while accounting for electricity costs, cold starts, and unmet demand, with terminal inventory and carbon penalties included where applicable. First, the scheduling problem is studied using grid electricity only, with electricity price as the main uncertain input. A deterministic Single-Execution framework provides the baseline, while rolling-horizon optimization (RHO) updates electricity-price forecasts and re-optimizes as operation proceeds. Across the grid-only case studies, RHO achieved higher realized profit on average than Single-Execution, demonstrating the potential value of incorporating updated information into scheduling decisions. Building on this foundation, the framework is extended to include on-site wind and solar generation, renewable forecasting, and hydrogen carbon intensity (CI). This allows electricity prices, renewable availability, and grid CI to be considered together when determining plant operation. Renewable integration reduced grid dependence and improved the ability of the plant to satisfy hydrogen-CI limits, particularly under high grid-CI conditions. Forecast updating also improved performance when new information led to better implemented decisions, although repeated re-optimization did not always result in a better overall schedule. Finally, two-stage stochastic programming is incorporated within RHO to form RHO+TSP, allowing multiple possible electricity-price, wind, and solar trajectories to be considered within each optimization. Single-Execution, RHO, and RHO+TSP are then compared under realized conditions and across a common external bank of 1,000 scenarios. Single-Execution and RHO both performed strongly, while RHO+TSP performed well in some cases but was less consistent and substantially more computationally expensive. RHO provided the most consistent balance among adaptability, economic performance, operational reliability, and computational effort. Overall, this work provides a modular framework for examining how forecast updating, renewable availability, CI constraints, hydrogen storage, realized renewable deviations, and explicit uncertainty interact in grid-connected green hydrogen scheduling. Future work could incorporate realized renewable deviations directly into the rolling optimization framework, evaluate longer operating periods, and explore integrated design and scheduling under uncertainty.","abstract_has_math":false,"creators":["Hamed, Lina"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":"Chemical Engineering","school":null,"contributors":[],"advisors":["Swartz, Christopher","Dalle Ave, Giancarlo"],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-18T19:37:28Z","subjects":["Green Hydrogen","PEM Electrolyzers","Scheduling Optimization","Stochastic Programming"],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://doi.org/10.71548/2406"],"render_values":[{"text":"https://doi.org/10.71548/2406","href":"https://doi.org/10.71548/2406","code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/11375/33566","outbound_label":"Handle","outbound_source":"dc:identifier.uri","record_url":"https://hdl.handle.net/11375/33566","record_label":"Repository record","record_kind":"handle","record_source":"dc:identifier.uri","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://macsphere.mcmaster.ca/server/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Amacsphere.mcmaster.ca%3A11375%2F33566","prefix":"dim"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Swartz, Christopher","Dalle Ave, Giancarlo"]},{"key":"dc:contributor.department","label":"Department","values":["Chemical Engineering"]},{"key":"dc:creator","label":"Author","values":["Hamed, Lina"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-09-18T14:51:55Z"]},{"key":"dc:date.issued","label":"Date","values":["2026"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Green Hydrogen","PEM Electrolyzers","Scheduling Optimization","Stochastic Programming"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/11375/33566","https://doi.org/10.71548/2406"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Green hydrogen produced through proton exchange membrane (PEM) electrolysis can support decarbonization, but its economic and environmental performance depends on interacting factors including electricity prices, the timing of electricity consumption, renewable availability, and grid carbon intensity. This thesis develops an adaptive and uncertainty-aware scheduling framework for a grid-connected multi-electrolyzer plant with hydrogen compression, storage, and on-site renewable generation. The objective maximizes profit while accounting for electricity costs, cold starts, and unmet demand, with terminal inventory and carbon penalties included where applicable. First, the scheduling problem is studied using grid electricity only, with electricity price as the main uncertain input. A deterministic Single-Execution framework provides the baseline, while rolling-horizon optimization (RHO) updates electricity-price forecasts and re-optimizes as operation proceeds. Across the grid-only case studies, RHO achieved higher realized profit on average than Single-Execution, demonstrating the potential value of incorporating updated information into scheduling decisions. Building on this foundation, the framework is extended to include on-site wind and solar generation, renewable forecasting, and hydrogen carbon intensity (CI). This allows electricity prices, renewable availability, and grid CI to be considered together when determining plant operation. Renewable integration reduced grid dependence and improved the ability of the plant to satisfy hydrogen-CI limits, particularly under high grid-CI conditions. Forecast updating also improved performance when new information led to better implemented decisions, although repeated re-optimization did not always result in a better overall schedule. Finally, two-stage stochastic programming is incorporated within RHO to form RHO+TSP, allowing multiple possible electricity-price, wind, and solar trajectories to be considered within each optimization. Single-Execution, RHO, and RHO+TSP are then compared under realized conditions and across a common external bank of 1,000 scenarios. Single-Execution and RHO both performed strongly, while RHO+TSP performed well in some cases but was less consistent and substantially more computationally expensive. RHO provided the most consistent balance among adaptability, economic performance, operational reliability, and computational effort. Overall, this work provides a modular framework for examining how forecast updating, renewable availability, CI constraints, hydrogen storage, realized renewable deviations, and explicit uncertainty interact in grid-connected green hydrogen scheduling. Future work could incorporate realized renewable deviations directly into the rolling optimization framework, evaluate longer operating periods, and explore integrated design and scheduling under uncertainty."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["Master of Applied Science (MASc)"]},{"key":"dc:title","label":"Title","values":["Adaptive and Uncertainty-Aware Optimization of Green Hydrogen Production via Grid-Connected PEM Electrolyzer Systems"]},{"key":"others:access_status","label":"Others Access Status","values":["embargo"]}]}],"canonical_facts":{"dc:contributor.advisor":["Swartz, Christopher","Dalle Ave, Giancarlo"],"dc:contributor.department":["Chemical Engineering"],"dc:creator":["Hamed, Lina"],"dc:date.accessioned":["2026-09-18T14:51:55Z"],"dc:date.issued":["2026"],"dc:description.abstract":["Green hydrogen produced through proton exchange membrane (PEM) electrolysis can support decarbonization, but its economic and environmental performance depends on interacting factors including electricity prices, the timing of electricity consumption, renewable availability, and grid carbon intensity. This thesis develops an adaptive and uncertainty-aware scheduling framework for a grid-connected multi-electrolyzer plant with hydrogen compression, storage, and on-site renewable generation. The objective maximizes profit while accounting for electricity costs, cold starts, and unmet demand, with terminal inventory and carbon penalties included where applicable. First, the scheduling problem is studied using grid electricity only, with electricity price as the main uncertain input. A deterministic Single-Execution framework provides the baseline, while rolling-horizon optimization (RHO) updates electricity-price forecasts and re-optimizes as operation proceeds. Across the grid-only case studies, RHO achieved higher realized profit on average than Single-Execution, demonstrating the potential value of incorporating updated information into scheduling decisions. Building on this foundation, the framework is extended to include on-site wind and solar generation, renewable forecasting, and hydrogen carbon intensity (CI). This allows electricity prices, renewable availability, and grid CI to be considered together when determining plant operation. Renewable integration reduced grid dependence and improved the ability of the plant to satisfy hydrogen-CI limits, particularly under high grid-CI conditions. Forecast updating also improved performance when new information led to better implemented decisions, although repeated re-optimization did not always result in a better overall schedule. Finally, two-stage stochastic programming is incorporated within RHO to form RHO+TSP, allowing multiple possible electricity-price, wind, and solar trajectories to be considered within each optimization. Single-Execution, RHO, and RHO+TSP are then compared under realized conditions and across a common external bank of 1,000 scenarios. Single-Execution and RHO both performed strongly, while RHO+TSP performed well in some cases but was less consistent and substantially more computationally expensive. RHO provided the most consistent balance among adaptability, economic performance, operational reliability, and computational effort. Overall, this work provides a modular framework for examining how forecast updating, renewable availability, CI constraints, hydrogen storage, realized renewable deviations, and explicit uncertainty interact in grid-connected green hydrogen scheduling. Future work could incorporate realized renewable deviations directly into the rolling optimization framework, evaluate longer operating periods, and explore integrated design and scheduling under uncertainty."],"dc:description.degree":["Master of Applied Science (MASc)"],"dc:identifier.uri":["https://hdl.handle.net/11375/33566","https://doi.org/10.71548/2406"],"dc:language.iso":["en"],"dc:subject":["Green Hydrogen","PEM Electrolyzers","Scheduling Optimization","Stochastic Programming"],"dc:title":["Adaptive and Uncertainty-Aware Optimization of Green Hydrogen Production via Grid-Connected PEM Electrolyzer Systems"],"dc:type":["Thesis"],"others:access_status":["embargo"]},"updated_at":"2026-11-18T19:37:28Z"}