Honeybee Incorrect Heat Balance

Replies to This Discussion

Permalink Reply by Chris Mackey on July 31, 2017 at 7:36pm

Hi Vincent,

I apologize for such a late reply.  You sent this message while I was in the middle of a move to a new home and it somehow slipped through the cracks.

Your interpretation of the First Law of Thermodynamics as it relates to buildings is only partially correct.  You are right that energy cannot be created or destroyed and that all energy that enters a building (through people, lighting, equipment, solar gain, or a heating system) must eventually leave the building (typically through envelope conduction, infiltration, mechanical/natural ventilation, or a cooling system).

However, the critical word in the statement above is EVENTUALLY.  It is very rare that, in any given moment, the energy  entering a building will equal the energy leaving it because buildings exist in a constant state of flux.  The weather around the building is always changing and the occupancy within the building is also changing.  So buildings are pretty much always in a state of either heating up slightly or cooling down slightly.  However, over long periods of time (like a month), you will see that the energy that entered the building is pretty much equal to the energy that left the building.  A good way to visualize this is with a monthly energy balance chart as you see in this example file:

ydrashare.github.io/hydra/viewer?owner=chriswmackey&fork=hydra_2&id=Energy_Balance

I hope that helps explain what E+ is doing under the hood,

-Chris

Permalink Reply by Vincent Höfte on August 1, 2017 at 2:34am

Hi Chris,

No worries on the late reply. The state of flux of a building explains the differences in heat losses and gains clearly.

I found that using the 'HB_Construct Energy Balance' component is far more simple and workable than constructing my own Energy Balance using the outputs of the 'Run Energyplus' component. When constructing the heat balance, including the chart, I found some strange results however. 

My simulation uses an Amsterdam -.epw file, which is characterized by a low temperature (avr. 4°C) in Feb and a higher temperature in August (avr. 18°C). However, my Opague conduction and my ventilation both have a higher heat loss during the summer months. I find this very odd, since the ΔT is much lower in these periods. (see image). First, I thought it had to do with solar radiation, but since this is calculated separately in the balance I don't think that's the problem.

Do you have any suggestion on where my model (or I) could be wrong?

Permalink Reply by Chris Mackey on August 1, 2017 at 6:16am

Hi Vincent,

The beautiful thing about energy balances is that you can figure out what causes a given term by looking at the opposite side of the chart. So your opaque conduction is increasing in summer because your solar gain is going up. Normally, I'd expect to see more of that solar taken out by the cooling system instead of going through opaque conduction. However, I would infer that you have chosen a high cooling setpoint, meaning that the building is heating up way above the average outdoor temperature when the sun hits it. The high temperature difference between inside and outside is driving this opaque conduction. Another possible factor is that your opaque walls might see less sun in the summer (when the sun is high in the sky). The lack of sun beating down on the walls means heat can conduct through the envelope more easily.

Energy models take some time to understand and debug before you trust them. I'd take a look through each other free hourly outputs on a 3D chart to understand what's going on in the model.

Permalink Reply by Vincent Höfte on August 3, 2017 at 1:11am

Hi Chris,

I agree with you on the beauty of an energy balance, for sure!

I've spend some time analyzing my results, but I still don't find them very logical. Your suggestions on a higher conduction due to a higher delta T or lower surface temperature seem logical. But looking at the hourly data shows me that something is interfering with the conduction data.

What happens; my conduction values get higher during daytime, while the ΔT is lower during this period. At some day, the conduction value even gets positive whilst the outside temperature is 6°C and the indoor is 19°C. See the figure below, it shows the increase in conduction loss during daytime and shows the conduction gain on sunday.

Looking at the shape of the thermal storage, I have the assumption that this is what's causing bias in my conduction data. 

Moreover, I found that when I use the Construct Energy Balance component, the sum of all my values is always zero. This makes sense if the temperature would stay equal, but in some timesteps my temperature rises/drops whilst the Energy Balance total sum is still zero. 

All of this gives me the assumption that something in the calculation is incorrect, but I double-checked all my inputs and can't seem to find where I go wrong. Based on the graph I attached, do you have an idea? If required, I could also upload my (stripped) -.gh.

Permalink Reply by Chris Mackey on August 12, 2017 at 5:51am

Hi Vincent,

Most of what you are describing makes sense when you account for the fact that EnergyPlus is a transient simulation and is a bit more sophisticated than a steady state heat balance.  For example, it is still possible to get positive conduction when outside temperatures are colder than indoors because the sun can beat down on the exterior of the construction and the construction absorbs some of this solar energy.  This can result in higher outdoor surface temperatures than indoor air temperatures and a corresponding conduction of heat to the building interior.

The reason why your heat balance is always summing to 0 is that the "storage" term refers to energy that is going into heating up or cooling down the mass of the building (including the mass of constructions and the mass of the interior air volume). Just use the output from the Energy Balance component that does not include storage and you will get an hourly value for how much energy is heating/cooling the mass of the building.

In any case, I can promise you that the EnergyPlus energy balance calculation is happening correctly in your simulation and that this calculation has been validated many times.  There are important differences between steady state and transient energy balances that take time to understand and to calculate.  But, in the end, the transient calculations are a closer representation to the complex reality of what buildings actually experience and there is a lot to be gained by accounting for this complexity.

-Chris

Permalink Reply by Pablo Vega Navarro on October 16, 2018 at 10:29pm

Good Evening,

How did you put this chart to work, I'm using the HB_Construct Energy Balance and it is not giving the graph.