Ground temperature average way to high

Hi everyone
simulating the basement of a building only, I wonder why the average ground floor temperature is around 18 °C?
According to the EPW file Location I would expect something like an average of 8°C.
Even If I use the standard construction Set ({0} 0. 2013::ClimateZone5::Mass) the result is pretty much the same.

Has anybody a clue what I’m missing here?

HB Lwh 002_ground.gh (118.2 KB)

@Martin6,

How are you getting 8C, is it the average of the ambient air temperature?

If so, that approximation for ground temperature doesn’t include the heat transfer from the ground slab to the earth (which has it’s own complicated dynamics[1]). Depending on the amount of insulation in your floor construction, 18C doesn’t seem that unreasonable.

[1] Computing ground temperature under an insulated glasshouse - #3 by SaeranVasanthakumar

@SaeranVasanthakumar

Thank you, for the helpful Links! Now things are way clearer.

The 8C are an empirical value for an unconditioned basement temperature in that area. Also, since the ground temperatures there in a depth of 1m vary throughout the year in between 1C and 17C I expected something in the middle of that range.

But I still wonder how to understand the 18C. Is that much heat pumped into the ground steadily?

You can try setting all ground boundary conditions to ‘Adiabatic’ to get an idea of what the temperature is without heat transfer, and also see what happens as you change your ground construction insulation.

However even an ‘Adiabatic’ boundary condition won’t match the ground temperatures given in the EPW file since modelers assume the ground temperature will be warmed by heat loss from slab, and is thus hardcoded by setting building-adjacent ground temperatures to a higher temperature (i.e 2C less then average internal temperature for commercial buildings[1]). The ground temperature in the EPW file represents undisturbed ground.

[1] Ground Modelling

1 Like

@Martin6 ,

As Saeran said, undisturbed ground temperatures are very different than the temperature beneath a floor slab of a conditioned space. 18 C is usually pretty close to reality for the temperature beneath a conditioned space in a heating dominated climate.

If you are in the tropics or you have measurements from beneath a ground slab, you can plug in custom monthly temperatures like so:


HB Lwh 002_ground.gh (118.8 KB)

Thank you @SaeranVasanthakumar , @chris for your input to my question.

Quite something to process and understand here.

The background for my consideration of the ground temperatures was, because in case of this building there is a 1m deep strip foundation all around and it’s isolated with 160mm EPS. So I was playing around the idea if there is a possibility of simulating the “heat” transfer from 1m below and see eventually if it’s enough energy to keep the room above “hotter” than 0C or whatever effect it has?

But for the moment I’m happy if I get a full building simulation running fine, so I better save that for later.

There are several more sophisticated ground heat transfer models built into EnergyPlus but we haven’t exposed them as part of honeybee. In particular, the KIVA model makes an attempt at trying to approximate the ground heat transfer as does the Ground slab preprocessor application.

Bu, if you are insulating your slab anyway, the heat transfer through it is bound to be so small compared to the other load balance terms of the model that it’s really overkill to use these more sophisticated models for most cases.

@chris @SaeranVasanthakumar , I have to dig this up because in recent simulations I came across this topic again.

Reading the .err file of a couple of simulations I read this notice

   ** Warning ** GetHTSurfaceData: Surfaces with interface to Ground found but no "Ground Temperatures" were input.
   **   ~~~   ** Found first in surface=1ST_1_5CBDFFBC..FACE4
   **   ~~~   ** Defaults, constant throughout the year of (18.0) will be used.

Which is kind of strange because in the .epw file there are ground temperatures for 0.5m 2m 4m.

GROUND TEMPERATURES,3,.5,,,,4.94,2.03,1.24,1.93,5.86,10.25,14.39,17.39,18.26,16.86,13.47,9.19,2,,,,7.56,4.83,3.54,3.49,5.56,8.57,11.81,14.58,16.01,15.75,13.83,10.89,4,,,,9.26,7.15,5.85,5.45,6.22,7.97,10.14,12.28,13.72,14.06,13.24,11.52

They don’t seem to be taken. What am i missing here?

@Martin6

The ground temperatures from the EPW file are computed to represent undisturbed soil, not ground temperatures of soil adjacent to building underground constructions, and thus aren’t used by EP to set ground temperatures.

Saeran is correct.

When there isn’t a user-specific input for ground temperatures, EnergyPlus just assumes a constant temperature of 18C on the other side of ground-contact Faces. This is good enough for estimation of energy use in most conditioned buildings (hence why this isn’t a severe warning). However, if you are modeling a passive building, you might want to specify some monthly ground temperatures since these will probably be a little bit closer to the undisturbed ground temperatures rather than a constant value. Also, if you are in the tropics, the constant temperature on the other side of ground Faces is probably higher than 18 C and closer to the annual average.

You can specify monthly ground temperatures using the Site:GroundTemperature:BuildingSurface EnergyPlus object with the add_str_ input on the Model To OSM component:

https://bigladdersoftware.com/epx/docs/9-6/input-output-reference/group-location-climate-weather-file-access.html#sitegroundtemperaturebuildingsurface

2 Likes

Hope KIVA model can be added to Honeybee :sunflower:

1 Like