Abstract
This paper describes the implementation of an energy management system (EMS) onto a semi-virtual nearly zero-energy building emulator.
The emulator platform consists of a physical part for energy production (PV panels, micro-wind turbine), storage (electrical battery and hot water tank) and conversion (ground-source heat pump and electric heater) coupled to a virtual single-family house model run in the computer simulation program TRNSYS. Communication between the parts is handled with a LabVIEW-based program.
The EMS is a Model-predictive Controller (MPC) Matlab-based program which optimizes energy flows in the system based on a successive linear programming (SLP) approach. The EMS is connected to real-time energy pricing and weather forecasts on the Internet and can instruct the emulator how to best utilize locally produced and imported energy.
In this paper, the performance of the EMS is evaluated by analyzing data from a one-week test period. We chose a representative week from April 2015 with varied energy pricing and weather conditions for this purpose. The actual test was done in June 2018, and hence we replaced the real energy production components with virtual ones to emulate the selected week’s energy generation from sun and wind. For the same reason, we replaced real-time energy pricing data with recorded ones and generated a weather forecast based on actual weather data from the representative week.
The measurement results indicate that the EMS improved the overall electricity use of the emulator building. It anticipated upcoming price increases by charging the battery and optimized self-consumption by covering the electricity demand with local production during more expensive morning hours and charging the battery with cheap imports during afternoons. The EMS also utilized the hot water tank to improve energy flexibility by allowing the bottom part to cool down during periods of low local generation and high space heating demand, and by heating up the tank during sunny and warm days.
The emulator platform consists of a physical part for energy production (PV panels, micro-wind turbine), storage (electrical battery and hot water tank) and conversion (ground-source heat pump and electric heater) coupled to a virtual single-family house model run in the computer simulation program TRNSYS. Communication between the parts is handled with a LabVIEW-based program.
The EMS is a Model-predictive Controller (MPC) Matlab-based program which optimizes energy flows in the system based on a successive linear programming (SLP) approach. The EMS is connected to real-time energy pricing and weather forecasts on the Internet and can instruct the emulator how to best utilize locally produced and imported energy.
In this paper, the performance of the EMS is evaluated by analyzing data from a one-week test period. We chose a representative week from April 2015 with varied energy pricing and weather conditions for this purpose. The actual test was done in June 2018, and hence we replaced the real energy production components with virtual ones to emulate the selected week’s energy generation from sun and wind. For the same reason, we replaced real-time energy pricing data with recorded ones and generated a weather forecast based on actual weather data from the representative week.
The measurement results indicate that the EMS improved the overall electricity use of the emulator building. It anticipated upcoming price increases by charging the battery and optimized self-consumption by covering the electricity demand with local production during more expensive morning hours and charging the battery with cheap imports during afternoons. The EMS also utilized the hot water tank to improve energy flexibility by allowing the bottom part to cool down during periods of low local generation and high space heating demand, and by heating up the tank during sunny and warm days.
| Original language | English |
|---|---|
| Publication status | Published - 2019 |
| MoE publication type | Not Eligible |
| Event | 9th International Conference on Sustainable Development in the Building and Environment, SuDBE2019 - University of Reading and University of Cambridge, United Kingdom Duration: 22 Jul 2018 → 28 Jul 2019 Conference number: 9 http://www.sudbeconference.com/ |
Conference
| Conference | 9th International Conference on Sustainable Development in the Building and Environment, SuDBE2019 |
|---|---|
| Abbreviated title | SuDBE |
| Country/Territory | United Kingdom |
| Period | 22/07/18 → 28/07/19 |
| Internet address |
Keywords
- nearly zero-energy buildings, energy management, renewables, building emulation
