A model that computes the per-unit process energy consumption, energy intensity, CO2 emission, and CO2 intensity of water treatment plants is developed. This model is used to estimate the total energy consumption of six water treatment plants in Seoul Metropolitan City (SMC), which is comprised 80–85% for finished water pumping, 6–10% for ozone disinfection, 2–4% for rapid mixing, and 1–3% for non-process loads. The model results are validated against actual data for 2020 and 2021. The net energy consumption considering renewable energy production and use is then calculated, and the corresponding level of CO2 emissions is predicted. Four scenarios based on the projected water requirements for the year 2045 were evaluated as follows: increased energy efficiency in finished water pumping (Scenario 1), increased renewable energy production in water treatment plants (Scenario 2), increased energy efficiency in raw water pumping (Scenario 3), and reduced water supply per capita (Scenario 4). Compared to a baseline do-nothing scenario (Scenario 0), the net energy consumption is reduced by 3.57%, 2.61%, 3.42%, and 4.67% for Scenarios 1–4, respectively. Scenario 4, which is a water-driven approach, is best for reducing CO2 emissions, while Scenario 1 and 3, which are energy-driven approaches, are more effective at reducing CO2 intensity.