In this paper, we show positive result for introducing energy conservation measures after analyzing facility data acquired by Energy Management System (EMS) in an existing middle-sized factory. The factory produces around 400,000 gas meters a year. To design new effective energy conservation measures, it was necessary to analyze room condition data and energy usage data. Therefore, real-time analysis of energy conservation was implemented by installing EMS for online energy data acquisition of electricity, high-pressure air volume, and room temperature. Based on energy consumption feature in the factory, energy conservation measures were created by 2 points of view. One is air volume for electricity reduction and another is optimization of HVAC for LPG reduction. For reduction of air, an air compressor was replaced with a new model based on a result of EMS data analysis. The replacement reduced electricity by around 50% at a peak time. For reduction of HVAC, chiller operation was improved. An analyzed each chiller's efficiency was different from a supposition of the administrator. To correct the operation smoothly, a chiller operation guidance system was developed using a data-driven modeling technology. The system predicted the heat load and generated chiller operation guidance. The operators set schedules in the chiller controller following the guidance. Changing chiller operation based on the result, chiller efficiency was improved by 8%. We achieved 36% CO₂ reduction in the period from 2011 to 2015 in the factory.