Calibrating a building energy model to hourly values

asked 2018-01-15 08:33:25 -0500

Cristina
95 ●4

updated 2018-01-15 10:06:41 -0500

I am trying to calibrate a passive house energy model using hourly measured data as input (interior and exterior temperature, lightings, electric equipment, infiltration). However, when I compare the simulated results with the monitoring data (hourly values for energy consumption) the variation is quite high. When I compare the monthly values, the variation is acceptable. Has anyone managed to calibrate a building energy model to hourly values? What are the input parameters that I should pay attention to?

I mention that the house has mechanical ventilation unit with heat recovery, air to water heat pump for heating and cooling and a buffer for hot water storage, solar collector for DHW and a boiler for storage. The heating and cooling is distributed through fan convecters installed in the ceiling.
The main objects I used for the systems in my model are: AirTerminal:SingleDuct:VAV:NoReheat, ZoneHVAC:PackagedTerminalAirConditioner, Coil:Cooling:DX:SingleSpeed and Coil:Heating:Water, HeatExchanger:AirToAir:FlatPlate For the heat pump I used Coil:WaterHeatingAirToWaterHeatPump:Pumped with WaterHeater:HeatPump:PumpedCondenser and WaterHeater:Mixed.

edit retag flag offensive close merge delete

add a comment

answered 2018-01-15 18:28:25 -0500

Anna Osborne Brannon

2917 ●7 ●23 http://www.atelierten.com/

It depends on what end uses you are seeing that differ significantly between the hourly and monthly results. For an hourly simulation, I'd start with schedules of operation and weather data. Once you have base loads calibrated (i.e. lighting and equipment) you can get a better idea of what other variables that these influence (i.e. heating and cooling) are off and by how much.

There are a lot of resources and guidance on how to approach model calibration, and there are some Q&A on this site on cloud simulation using OpenStudio that may be useful. Here is a paper on it as well: https://www.nrel.gov/docs/fy14osti/61...

From that paper, here are some common adjustments for model calibration:

image description